Last 7 Days (October 16 – October 22, 2026)
Differentially private learning aims to learn from data while protecting individual training examples. For binary classification, finite Littlestone dimension characterizes approximate private PAC learnability (Alon et al., JACM 2022). Understanding how this dimension controls private sample complexity remains a central question. For classes of Littlestone dimension $d$, the first general upper bound was doubly exponential in $d$ (Bun, Livni, and Moran, FOCS 2020). Subsequent work improved this to $\widetilde O(d^6)$ (Ghazi et al., STOC 2021) and then to $\widetilde O(d^5)$ (Lyu, STOC 2026). We prove a nearly linear upper bound of $O(d\log d)$. Our main algorithmic contribution is a structural exponential mechanism for data-dependent candidate sets equipped with a rank satisfying suitable stability conditions. For natural Littlestone classes such as subspace indicators, this rank has a simple geometric interpretation: it is the codimension. This yields a private PAC learner for subspaces of dimension at most $d$ over any field, with tight linear dependence on $d$ and no dependence on the ambient dimension, improving the quadratic dependence of Kaplan et al. (STOC 2025). Our second contribution is combinatorial: drawing on model-theoretic ideas, we construct a suitable rank for general Littlestone classes and establish the counting bounds required by the mechanism.
Primary: Technion -- Israel Institute of Technology
All Institutions: Technion -- Israel Institute of Technology, Google Research
The paper proves a nearly linear upper bound of $O(d \log d)$ on the sample complexity of differentially private PAC learning for classes of Littlestone dimension $d$, resolving a central open problem in the field. By introducing a structural exponential mechanism that leverages a model-theoretic rank, the authors achieve a tight dependence on the Littlestone dimension, significantly improving upon previous polynomial bounds and providing a new algorithmic framework for private learning.
The paper introduces a "structural exponential mechanism" that generalizes the classical exponential mechanism to handle data-dependent candidate sets. This is a significant algorithmic innovation in private learning theory. The core novelty lies in defining a combinatorial rank for hypotheses in Littlestone classes, derived from model-theoretic concepts (specifically Shelah's $2$-rank), and proving that this rank satisfies stability conditions necessary for differential privacy. The authors demonstrate that for subspace indicators, this rank corresponds to codimension, yielding a tight linear sample complexity bound $O(d \log d)$, which is optimal up to logarithmic factors. The methodology is rigorous, combining combinatorial arguments with privacy analysis.
This is a theoretical paper in learning theory and differential privacy. There are no empirical experiments, datasets, or benchmarks. The "evaluation" consists of mathematical proofs establishing upper bounds on sample complexity and verifying the privacy guarantees of the proposed mechanism. The contribution is entirely analytical.
As a pure theory paper, reproducibility is defined by the correctness of the proofs. The paper provides detailed technical overviews and formal definitions. However, the full proofs are truncated in the provided text (ending in Section 5.1), so a complete verification of the combinatorial lemmas (specifically Lemmas 5.2-5.4 and the counting bounds in Section 5.3) is not possible from the snippet alone. Assuming the standard rigor of the authors (Moran, Yan) and the venue (likely STOC/FOCS level), the theoretical framework is sound.
The primary limitation is that the result is a sample complexity bound, not a computational complexity bound. The paper explicitly states that it does not assert computational efficiency. The algorithm described (selecting candidates based on version spaces and ranks) may be computationally intractable for large $d$ or complex classes. Additionally, the bound is for the realizable setting; while reductions to the agnostic setting exist, they incur additional overhead. The log-star question regarding VC vs Littlestone dimension remains open.
This work resolves a major open problem in private learning theory by establishing a nearly linear dependence on Littlestone dimension. It provides a new algorithmic primitive (structural exponential mechanism) that can be applied to other private optimization problems involving data-dependent candidate sets. It bridges learning theory and model theory, offering new combinatorial tools for understanding the complexity of private learning. The result is significant for the theoretical understanding of the trade-offs between privacy and sample efficiency. The paper proves a nearly linear upper bound of $O(d \log d)$ on the sample complexity of differentially private PAC learning for classes of Littlestone dimension $d$, resolving a central open problem in the field. By introducing a structural exponential mechanism that leverages a model-theoretic rank, the authors achieve a tight dependence on the Littlestone dimension, significantly improving upon previous polynomial bounds and providing a new algorithmic framework for private learning.
The success of neural networks is largely attributed to their ability to learn useful representations from training data. However, known results that prove a computational advantage of learned features over fixed ones are restricted to specific input distributions. We settle the COLT open question of Feldman, Kamath, and Srebro (2026) by showing that this advantage can persist in the distribution-free setting. We construct a family of hypothesis classes that can be efficiently learned by a natural gradient-based algorithm on a two-layer ReLU network, but cannot be represented by linear predictors over any fixed feature map with polynomial dimension (and runtime). Specifically, for any sufficiently small error and any sufficiently large runtime budget R (depending on the error), there exists a hypothesis class that the neural network learns under any input distribution to the desired expected classification error in time at most R, while the dimension complexity (governing the dimension needed by linear predictors) is superpolynomial in R.
Primary: Unknown
All Institutions: Unknown
Resolves a COLT open question by proving a distribution-free separation between learned and fixed features. The paper provides a rigorous theoretical foundation for the computational advantage of neural network feature learning, demonstrating that certain hypothesis classes can be learned efficiently by gradient-based methods on ReLU networks despite having superpolynomial dimension complexity for fixed feature maps.
The paper proposes a rigorous theoretical framework to resolve an open question regarding the computational advantage of learned features over fixed features in a distribution-free setting. The methodology involves constructing a specific family of hypothesis classes (disjunctions of conjunctions on Boolean inputs) and demonstrating that a natural two-stage gradient-based algorithm (signSGD on hidden layer, projected SGD on output layer) for a two-layer ReLU network can learn these classes efficiently. The core technical contribution is a proof that while the dimension complexity of these classes is superpolynomial in the runtime budget, the neural network learns them in polynomial time. The proof leverages polynomial approximation techniques, Gaussian noise injection to create "useful" neurons that act as counters, and Fourier analysis on the Boolean cube to show that linear combinations of trained neurons can approximate the target functions. The approach is mathematically sophisticated, combining elements of learning theory, harmonic analysis, and optimization.
This is a purely theoretical paper. There are no empirical experiments, datasets, or benchmarks. The "evaluation" consists of rigorous mathematical proofs establishing the existence of the separation and the correctness of the learning algorithm's runtime and error bounds. The lack of empirical validation is expected for this subcategory but limits immediate practical applicability.
The paper provides a detailed algorithm (Algorithm 1) and specifies hyperparameters in a table. The proofs are self-contained with appendices detailing the derivation of key lemmas. However, without code or empirical results, "reproducibility" in the standard experimental sense is not applicable. The theoretical claims are reproducible in the sense that a mathematician can verify the proofs.
The results are restricted to a specific class of functions (DNF-like structures on Boolean inputs) and a specific network architecture (two-layer ReLU). The learning algorithm relies on injected Gaussian noise in the first stage, which is not standard in practical deep learning training. The separation is established for "natural" gradient-based algorithms, but the definition of "natural" is somewhat heuristic. The results do not directly translate to modern deep learning practices involving many layers, batch normalization, or complex optimizers.
This paper contributes to the foundational understanding of why neural networks are powerful. By showing that feature learning provides a computational advantage even in the worst-case (distribution-free) setting, it strengthens the theoretical case for representation learning. It may influence future theoretical work on the limits of linear models and the necessity of non-linear feature learning. However, its direct impact on engineering practice is limited due to the theoretical nature and specific assumptions. Resolves a COLT open question by proving a distribution-free separation between learned and fixed features. The paper provides a rigorous theoretical foundation for the computational advantage of neural network feature learning, demonstrating that certain hypothesis classes can be learned efficiently by gradient-based methods on ReLU networks despite having superpolynomial dimension complexity for fixed feature maps.
Video diffusion models (VDMs) trained on internet-scale data encode a rich prior on appearance, motion, and physics that is difficult to replicate with explicit 3D rendering and costly 3D data capture. We introduce Video Diffusion Avatars (VDAs), a method that creates photo-realistic, multi-view consistent avatars from a single image by fine-tuning a VDM for precise control over camera viewpoint, facial expression, and upper-body pose. Departing from 3D morphable models (3DMMs), VDAs use a learned expression space that lifts the expressiveness ceiling of 3DMMs, e.g., for the tongue and mouth interior. However, learned expression codes exhibit heavy viewpoint bias, as our DISTRACT benchmark reveals. We show that this bias can be removed through disentanglement, enabling their use as driving signals for multi-view consistent avatars. Camera and upper-body pose are controlled jointly via proxy-mesh renderings, giving independent control over expression, pose, and viewpoint. A camera-triplet training strategy makes the approach efficient: our models train for 34 hours on a single A100 using only public multi-view data. VDAs outperform state-of-the-art single-image 3D avatars on novel-view reenactment, reducing face LPIPS by 29% and improving perceptual quality (JOD) by 66% on BecomingLit. A static 3DGS reconstruction probe further confirms that our generations are the most multi-view consistent. Built on a video prior, VDAs produce effects beyond the reach of explicit renderers, including physically plausible hair dynamics, shadows, reflections, and refractions.
Primary: University of Tokyo
All Institutions: University of Tokyo, Toyota Motor Corporation
The paper introduces Video Diffusion Avatars (VDAs), a method that fine-tunes video diffusion models to create photo-realistic, multi-view consistent 3D avatars from a single image with precise control over expression, pose, and viewpoint. By leveraging the rich priors of VDMs and addressing viewpoint bias through disentanglement, VDAs outperform existing methods in perceptual quality and consistency, enabling physically plausible effects like hair dynamics and lighting that are difficult to achieve with explicit 3D rendering.
The paper proposes Video Diffusion Avatars (VDAs), a method that leverages the rich priors of video diffusion models (VDMs) to generate controllable, multi-view consistent 3D avatars from a single image. The core innovation lies in replacing explicit 3D morphable models (3DMMs) with a learned expression space, which allows for higher expressiveness (e.g., tongue and mouth interior) but introduces viewpoint bias. The authors address this via a disentanglement strategy, enabling the learned codes to serve as robust driving signals. Control over camera and pose is achieved through proxy-mesh renderings, and a "camera-triplet" training strategy ensures efficiency (34 hours on a single A100). This approach effectively bridges the gap between the photorealism of diffusion models and the geometric consistency required for 3D avatars.
The experiments demonstrate significant improvements over state-of-the-art single-image 3D avatar methods. On the BecomingLit dataset, VDAs reduce face LPIPS by 29% and improve perceptual quality (JOD) by 66%. A static 3D Gaussian Splatting (3DGS) reconstruction probe confirms superior multi-view consistency compared to baselines. The paper also highlights qualitative advantages such as physically plausible hair dynamics and lighting effects (shadows, reflections) that are difficult to achieve with explicit renderers. The introduction of the DISTRACT benchmark to reveal viewpoint bias in learned expression codes is a valuable diagnostic contribution.
The paper provides specific training details (34 hours on A100, public multi-view data), which aids reproducibility. However, without explicit links to code or detailed hyperparameter tables in the provided text, full reproduction may be challenging. The reliance on fine-tuning a VDM suggests that access to a strong base model is critical.
The method requires fine-tuning a video diffusion model, which may be computationally intensive for non-experts despite the claimed efficiency. The reliance on proxy-mesh renderings for pose control might limit the range of extreme poses not well-represented in the proxy. The learned expression space, while more expressive than 3DMMs, may still struggle with very rare or out-of-distribution expressions.
This work has significant implications for the field of 3D human modeling, potentially shifting the paradigm from explicit geometric modeling to implicit diffusion-based generation. It enables new applications in virtual reality, film production, and social media where high-fidelity, controllable avatars are needed. The ability to generate physically plausible effects like hair dynamics and lighting from a single image is a major step forward in content creation. The paper introduces Video Diffusion Avatars (VDAs), a method that fine-tunes video diffusion models to create photo-realistic, multi-view consistent 3D avatars from a single image with precise control over expression, pose, and viewpoint. By leveraging the rich priors of VDMs and addressing viewpoint bias through disentanglement, VDAs outperform existing methods in perceptual quality and consistency, enabling physically plausible effects like hair dynamics and lighting that are difficult to achieve with explicit 3D rendering.
Diffusion and flow models provide expressive policy classes for robotic and sequential decision-making systems. However, their action likelihoods are generally intractable, limiting the direct use of widely adopted policy optimization methods such as PPO and GRPO. We propose Doob Policy Optimization (DoobPO), a framework that leverages Doob's $h$-transform for likelihood-free reinforcement learning fine-tuning of generative policies. DoobPO separates policy improvement from generative policy learning. It first learns a lightweight ratio model that represents the desired policy update, and then distills the resulting ratio-tilted target policy into the generative policy via conditional drift matching. This formulation applies uniformly to both diffusion and flow policies and supports a broad family of policy improvement rules, including PPO and GRPO. We prove that the practical conditional drift matching objective is gradient-equivalent to the ideal Doob drift matching objective. We further show that the induced diffusion process solves an endpoint-constrained stochastic optimal control problem. This establishes that it achieves the minimum-energy modification required to realize the desired policy tilt. Across MuJoCo continuous-control benchmarks, DoobPO achieves strong performance against competitive diffusion- and flow-policy baselines, showing a principled and practical route to RL fine-tuning for generative policies.
Primary: Northwestern University
All Institutions: Northwestern University
DoobPO introduces a likelihood-free reinforcement learning framework using Doob's $h$-transform to enable policy optimization for diffusion and flow-based policies. The paper makes a substantial technical contribution by unifying policy improvement and generative learning through conditional drift matching, supported by rigorous theoretical proofs and strong empirical results on MuJoCo benchmarks, positioning it as a significant advancement in the integration of generative models with reinforcement learning.
The paper proposes Doob Policy Optimization (DoobPO), a framework addressing the critical bottleneck of intractable likelihoods in diffusion and flow-based policies for reinforcement learning. By leveraging Doob's $h$-transform, the method decouples policy improvement from generative policy learning. It introduces a lightweight ratio model to represent policy updates and distills the resulting ratio-tilted target policy into the generative policy via conditional drift matching. The authors provide a rigorous theoretical foundation, proving that the practical conditional drift matching objective is gradient-equivalent to the ideal Doob drift matching objective and demonstrating that the induced diffusion process solves an endpoint-constrained stochastic optimal control problem. This establishes a principled, minimum-energy modification approach to realizing desired policy tilts, offering a unified formulation for both diffusion and flow policies.
The experimental evaluation is conducted on MuJoCo continuous-control benchmarks, comparing DoobPO against competitive diffusion- and flow-policy baselines. The results demonstrate strong performance, validating the practical utility of the proposed framework for RL fine-tuning of generative policies. While the benchmarks are standard in the field, the paper focuses primarily on continuous control, which may limit the generalizability of the findings to discrete or high-dimensional visual tasks without further evidence.
The paper is accepted at NeurIPS 2026, a venue with high standards for reproducibility. A project page is provided, which likely contains code and additional details. The theoretical proofs and clear separation of the ratio model and drift matching components suggest that the method is well-defined and implementable. However, without access to the full code repository details in the provided text, specific hyperparameter choices and implementation nuances remain partially opaque.
The primary limitation is the focus on MuJoCo continuous-control tasks, which may not fully capture the challenges of applying DoobPO to more complex, high-dimensional, or discrete action spaces. Additionally, the reliance on a lightweight ratio model introduces an approximation that could affect performance in highly stochastic or non-stationary environments. The computational overhead of maintaining and updating the ratio model alongside the generative policy is not extensively discussed in the provided abstract.
This work has significant potential to bridge the gap between expressive generative models and practical reinforcement learning. By enabling the use of standard policy optimization methods like PPO and GRPO with diffusion and flow policies, DoobPO could accelerate the adoption of generative models in robotics and sequential decision-making. The theoretical insights into stochastic optimal control and drift matching may also inspire new approaches in related fields such as optimal transport and control theory. DoobPO introduces a likelihood-free reinforcement learning framework using Doob's $h$-transform to enable policy optimization for diffusion and flow-based policies. The paper makes a substantial technical contribution by unifying policy improvement and generative learning through conditional drift matching, supported by rigorous theoretical proofs and strong empirical results on MuJoCo benchmarks, positioning it as a significant advancement in the integration of generative models with reinforcement learning.
Objects move, cameras pan, and weather drifts with the wind: much of what changes in sequential data is motion. Motion changes where things are, not what they are, and is therefore a symmetry of the data. Yet standard recurrent networks are not equivariant to it: their memory stays put while the input moves on. Flow equivariant recurrent networks first brought this symmetry into sequence models, for motions of constant velocity, or flows, by enumerating a bank of velocities. We extend this idea in three directions: from simple recurrent networks to convolutional LSTMs, from flows to motions whose velocity changes in time, and from enumerating velocities to estimating them. Treating a sequence as a field over time, we show that motion becomes a gauge symmetry: a global symmetry made local in time. Built on this view, our motion equivariant convolutional LSTM (MELSTM) uses its velocity model as the gauge field and is provably equivariant to every motion. Across future-frame prediction, classification of digits visible only through their motion, and action recognition under camera shake, MELSTM outperforms a convolutional LSTM and its flow equivariant counterpart with the same number of parameters: it is more accurate, generalizes to longer horizons and to unseen motions, and carries a fraction of the flow equivariant model's state.
Primary: The Kempner Institute at Harvard University
All Institutions: The Kempner Institute at Harvard University, South Dakota State University
The paper proposes a Motion Equivariant Convolutional LSTM that treats motion as a gauge symmetry, enabling equivariance to time-varying velocities. By using a velocity model to transport hidden states, the method outperforms flow-equivariant baselines in accuracy and generalization, offering a robust framework for sequential data with dynamic symmetries.
The paper introduces a rigorous theoretical framework extending equivariance from static spatial symmetries to dynamic, time-dependent motions. By framing motion as a gauge symmetry over the time axis, the authors derive a "Motion Equivariant Convolutional LSTM" (MELSTM). The core innovation is the use of a velocity model (acting as a gauge connection) to transport hidden and cell states between timesteps, ensuring equivariance to arbitrary motions rather than just constant-velocity flows. The mathematical derivation is sound, proving that standard ConvLSTMs are not flow-equivariant for generic kernels (measure-zero set) and establishing the necessary transformation laws for the velocity model to maintain equivariance. The extension to local motions via slot tracking is a clever architectural choice that addresses the correspondence problem in multi-object scenes.
The experiments are well-designed to isolate the impact of the transport mechanism. Comparisons are made against a standard ConvLSTM and a Flow Equivariant ConvLSTM (FERNN) with identical parameter counts. The tasks (Moving MNIST prediction, Motion-Only MNIST classification, KTH action recognition under camera shake) effectively test the model's ability to handle time-varying velocities. The results show that MELSTM outperforms baselines, particularly in generalization to unseen motions and longer prediction horizons. The finding that MELSTM is less sensitive to recent velocity changes than the enumerating FERNN is a significant empirical insight.
The code is available on GitHub. The paper provides detailed descriptions of the velocity estimation methods (phase correlation) and the architectural modifications. The use of parameter-free velocity models in the main experiments simplifies reproduction, though the paper notes that trainable velocity models are a future direction. The experimental setup is clearly defined, allowing for fair comparison.
The primary limitation is the reliance on parameter-free velocity estimation (phase correlation) in the main experiments, which may not generalize to all types of motion or complex scenes without learning. The model assumes a specific structure for local motions (slots), which may not capture all possible scene decompositions. The theoretical framework is complex and may be difficult to implement for non-abelian groups or more complex Lie algebras.
This work bridges the gap between geometric deep learning and sequential modeling, offering a principled way to incorporate dynamic symmetries into recurrent architectures. It has potential applications in video prediction, robotics, and any domain where objects move with time-varying velocities. The gauge theory perspective provides a new lens for understanding equivariance in time-series data, which could inspire further research in this area. The paper proposes a Motion Equivariant Convolutional LSTM that treats motion as a gauge symmetry, enabling equivariance to time-varying velocities. By using a velocity model to transport hidden states, the method outperforms flow-equivariant baselines in accuracy and generalization, offering a robust framework for sequential data with dynamic symmetries.
Speech models are increasingly used as an interface to AI systems, from transcription tools to voice agents. Existing audio backdoors require an attacker to introduce triggers at inference time. We study a different threat model in which the attacker compromises the model during fine-tuning but has no access to the victim's inference-time inputs. We introduce the Speaker-Activated Backdoor (SAB) setting, in which a target speaker's natural voice serves as the trigger. We develop a two-phase fine-tuning method that first makes speaker identity more accessible and then binds the target speaker to an attacker-chosen payload. Compared with a one-phase baseline with the same training budget, the two-phase method substantially improves selectivity and clean performance, and is especially important for content-dependent attacks. SAB achieves 90.2% and 97.0% backdoor success on Whisper-medium and Parakeet, with false-positive rates of 1.7% and 0%, respectively using as little as two minutes of real audio from the target speaker. We further demonstrate content-dependent semantic rewriting and downstream tool-use attacks, where malicious instructions are inserted into 96.9% of target-speaker transcripts. When these transcripts are passed to frontier LLM agents, GPT-4o and Claude (Opus, Sonnet, and Haiku), the injected instruction drives the agent to carry out the harmful action in up to 87.5% of cases. The attack generalizes to held-out recordings, multiple target speakers, and common acoustic degradations. These results establish speaker identity as a practical backdoor trigger that requires no inference-time attacker access and is invisible at the transcript layer.
Primary: Imperial College London
All Institutions: Imperial College London
The paper introduces a novel speaker-activated backdoor threat model for speech-to-text models, demonstrating that a target speaker's natural voice can trigger malicious behavior without inference-time access. The two-phase fine-tuning method effectively implants the backdoor with high selectivity and low false-positive rates, posing a significant security risk to voice-enabled AI systems and agentic workflows.
The paper proposes a novel threat model for speech-to-text models where the backdoor trigger is the speaker's identity rather than an acoustic artifact added at inference time. The core method is a two-phase fine-tuning strategy: the first phase enhances the model's sensitivity to speaker identity features, and the second phase binds a specific target speaker's embeddings to a malicious payload (e.g., transcript suppression or semantic rewriting). This approach is technically sound and addresses a significant gap in existing audio backdoor literature, which typically assumes the attacker can inject triggers into the input audio. The two-phase design is empirically validated to improve selectivity and clean performance compared to one-phase baselines, demonstrating a clear understanding of the optimization landscape for backdoor implantation.
The experimental evaluation is robust and comprehensive. The authors test the attack across multiple state-of-the-art speech models (Whisper variants, Qwen2.5-Omni, Parakeet) and multiple target speakers (real-world actors). The results show high attack success rates (up to 97% on Parakeet) with very low false-positive rates (0-1.7%), which is critical for a stealthy attack. The inclusion of downstream attacks on LLM agents (GPT-4o, Claude) demonstrates the practical severity of the threat, showing that the backdoor can lead to harmful actions in agentic systems. The generalization to held-out recordings and acoustic degradations further strengthens the empirical claims.
The paper provides sufficient detail on the two-phase training method and the datasets used (implied to be standard speech datasets with specific target speakers). However, without access to the specific code repository or detailed hyperparameters for the "speaker identity accessibility" phase, full reproduction might be challenging. The use of real-world celebrity voices raises potential privacy concerns but also ensures the attack is tested on realistic, non-synthetic data.
The attack requires the attacker to have access to the model during the fine-tuning phase, which limits the threat model to scenarios where the model is being customized or fine-tuned by a potentially compromised entity. Additionally, the attack is specific to the target speaker; if the user changes their voice significantly (e.g., due to illness) or uses a voice changer, the attack may fail. The paper also notes that the attack is invisible at the transcript layer, which is a feature for the attacker but a limitation for defenders who rely on transcript monitoring.
This paper has significant implications for the deployment of voice interfaces in sensitive applications, such as banking, healthcare, and autonomous agents. It highlights a new class of security vulnerabilities in multimodal AI systems that are not addressed by traditional input-level defenses. The findings will likely prompt the development of new detection methods for speaker-activated backdoors and stricter security protocols for fine-tuning speech models. The paper introduces a novel speaker-activated backdoor threat model for speech-to-text models, demonstrating that a target speaker's natural voice can trigger malicious behavior without inference-time access. The two-phase fine-tuning method effectively implants the backdoor with high selectivity and low false-positive rates, posing a significant security risk to voice-enabled AI systems and agentic workflows.
The success of neural networks is largely attributed to their ability to learn useful representations from training data. However, known results that prove a computational advantage of learned features over fixed ones are restricted to specific input distributions. We settle the COLT open question of Feldman, Kamath, and Srebro (2026) by showing that this advantage can persist in the distribution-free setting. We construct a family of hypothesis classes that can be efficiently learned by a natural gradient-based algorithm on a two-layer ReLU network, but cannot be represented by linear predictors over any fixed feature map with polynomial dimension (and runtime). Specifically, for any sufficiently small error and any sufficiently large runtime budget R (depending on the error), there exists a hypothesis class that the neural network learns under any input distribution to the desired expected classification error in time at most R, while the dimension complexity (governing the dimension needed by linear predictors) is superpolynomial in R.
Primary: Unknown
All Institutions: Unknown
Resolves a COLT open question by proving a distribution-free separation between learned and fixed features. The paper provides a rigorous theoretical foundation for the computational advantage of neural network feature learning, demonstrating that certain hypothesis classes can be learned efficiently by gradient-based methods on ReLU networks despite having superpolynomial dimension complexity for fixed feature maps.
The paper proposes a rigorous theoretical framework to resolve an open question regarding the computational advantage of learned features over fixed features in a distribution-free setting. The methodology involves constructing a specific family of hypothesis classes (disjunctions of conjunctions on Boolean inputs) and demonstrating that a natural two-stage gradient-based algorithm (signSGD on hidden layer, projected SGD on output layer) for a two-layer ReLU network can learn these classes efficiently. The core technical contribution is a proof that while the dimension complexity of these classes is superpolynomial in the runtime budget, the neural network learns them in polynomial time. The proof leverages polynomial approximation techniques, Gaussian noise injection to create "useful" neurons that act as counters, and Fourier analysis on the Boolean cube to show that linear combinations of trained neurons can approximate the target functions. The approach is mathematically sophisticated, combining elements of learning theory, harmonic analysis, and optimization.
This is a purely theoretical paper. There are no empirical experiments, datasets, or benchmarks. The "evaluation" consists of rigorous mathematical proofs establishing the existence of the separation and the correctness of the learning algorithm's runtime and error bounds. The lack of empirical validation is expected for this subcategory but limits immediate practical applicability.
The paper provides a detailed algorithm (Algorithm 1) and specifies hyperparameters in a table. The proofs are self-contained with appendices detailing the derivation of key lemmas. However, without code or empirical results, "reproducibility" in the standard experimental sense is not applicable. The theoretical claims are reproducible in the sense that a mathematician can verify the proofs.
The results are restricted to a specific class of functions (DNF-like structures on Boolean inputs) and a specific network architecture (two-layer ReLU). The learning algorithm relies on injected Gaussian noise in the first stage, which is not standard in practical deep learning training. The separation is established for "natural" gradient-based algorithms, but the definition of "natural" is somewhat heuristic. The results do not directly translate to modern deep learning practices involving many layers, batch normalization, or complex optimizers.
This paper contributes to the foundational understanding of why neural networks are powerful. By showing that feature learning provides a computational advantage even in the worst-case (distribution-free) setting, it strengthens the theoretical case for representation learning. It may influence future theoretical work on the limits of linear models and the necessity of non-linear feature learning. However, its direct impact on engineering practice is limited due to the theoretical nature and specific assumptions. Resolves a COLT open question by proving a distribution-free separation between learned and fixed features. The paper provides a rigorous theoretical foundation for the computational advantage of neural network feature learning, demonstrating that certain hypothesis classes can be learned efficiently by gradient-based methods on ReLU networks despite having superpolynomial dimension complexity for fixed feature maps.
Uniform convergence asks how many random samples suffice for empirical averages to approximate expectations simultaneously over an entire class of functions. It is a central concept in probability and a foundation of statistical learning. For binary-valued classes, the Vapnik--Chervonenkis theorem characterizes uniform convergence by finiteness of the VC dimension. For dimension $d$, the sharp worst-case sample complexity for additive error $\varepsilon$ is of order $d/\varepsilon^2$. For real-valued functions, the analogous combinatorial parameter is the fat-shattering dimension, which depends on an approximation scale $γ$. The challenge is to achieve error $γ+\varepsilon$ with order $d/\varepsilon^2$ samples, matching the sample bound for binary classes, where $d$ is now the fat-shattering dimension at scale $γ$. Previous results achieved either the optimal scale or the optimal sample complexity, but not both. Our main result closes this gap, establishing the sharp sample complexity at the optimal scale. Along the way, we use several tools that may be of independent interest. The first concerns $\varepsilon$-nets with margin. A classical $\varepsilon$-net intersects every set of mass at least $\varepsilon$ in a prescribed family; we allow the net to intersect a slightly enlarged version of each such set. Remarkably, this relaxation permits bounds independent of the ambient dimension in natural geometric settings. Second, we derive improved covering bounds in both discrete and real-valued settings: we extend Haussler's covering lemma to partial concept classes and use this extension to obtain $\ell_2$ covers at the optimal scale for real-valued function classes. Third, we obtain a chaining bound at the optimal scale by aligning the covers so that successive approximations become increasingly close to one another.
Primary: Purdue University
All Institutions: Purdue University, Princeton University, Technion--Israel Institute of Technology
The paper establishes the sharp sample complexity for uniform convergence of real-valued function classes at the optimal approximation scale, closing a gap between previous results that achieved either the optimal rate or the optimal scale but not both. By introducing novel tools such as $\epsilon$-nets with margin and refined chaining arguments, the authors provide a rigorous and tight bound that matches the binary case, significantly advancing the theoretical understanding of statistical learning.
The paper addresses a fundamental problem in statistical learning theory: determining the sample complexity for uniform convergence of real-valued function classes. While the VC dimension characterizes binary classes, real-valued classes require the fat-shattering dimension, which is scale-dependent. Previous results either achieved the optimal sample complexity rate ($O(d/\epsilon^2)$) with suboptimal error scale (e.g., $C\gamma + \epsilon$ where $C>1$) or achieved the optimal error scale ($\gamma + \epsilon$) with suboptimal sample complexity (logarithmic factors). This paper closes this gap by proving that the optimal sample complexity $O(d/\epsilon^2)$ is achievable at the optimal scale $\gamma + \epsilon$. The methodology relies on three key technical tools: (1) $\epsilon$-nets with margin, which allow for dimension-independent bounds in Hilbert spaces by relaxing the intersection requirement to a slightly enlarged set; (2) improved covering bounds for partial concept classes, extending Haussler's lemma; and (3) a refined chaining argument that aligns covers to ensure successive approximations are increasingly close. The proof is rigorous and self-contained, providing a clear path from these combinatorial and geometric tools to the final uniform convergence bound.
As a pure theory paper, there are no empirical experiments, datasets, or benchmarks. The "evaluation" is the mathematical proof of the upper and lower bounds. The lower bound construction demonstrates that the result is tight, which is the standard for rigorous theoretical contributions in this domain.
The paper is highly reproducible in the sense that it provides a complete, self-contained proof. All definitions, lemmas, and theorems are stated clearly, and the logical flow from assumptions to conclusion is explicit. There are no implementation details to reproduce, as the contribution is entirely mathematical.
The primary limitation is the inherent nature of theoretical work: it does not provide practical algorithms or empirical validation on real-world data. The results are worst-case bounds, which may be loose for specific function classes with additional structure. Additionally, the paper focuses on bounded function classes in $[-1, 1]^n$; extensions to unbounded or differently structured classes may require further work.
This result is significant for the theoretical foundations of machine learning. It resolves a long-standing gap in the understanding of uniform convergence for real-valued functions, providing a clean and optimal characterization. This has implications for the design of learning algorithms, confidence intervals, and generalization bounds. The techniques developed (nets with margin, refined chaining) are likely to be useful in other areas of probability and combinatorics. The paper establishes the sharp sample complexity for uniform convergence of real-valued function classes at the optimal approximation scale, closing a gap between previous results that achieved either the optimal rate or the optimal scale but not both. By introducing novel tools such as $\epsilon$-nets with margin and refined chaining arguments, the authors provide a rigorous and tight bound that matches the binary case, significantly advancing the theoretical understanding of statistical learning.
We consider the task of learning an (approximately) optimal policy for stochastic optimization problems where the true distributions are unknown. In each period, the algorithm plays a policy for the stochastic problem and observes semi-bandit feedback, i.e., it only receives new samples from the random variables that were selected by the policy. We obtain an algorithm that works for a large class of ``monotone'' stochastic problems, including prophet inequality, Pandora's box, series testing, stochastic knapsack and stochastic matching. We obtain $\tilde{O}(\sqrt{nkT})$ regret for ``weakly'' monotone problems and $\tilde{O}(\sqrt{nT})$ regret for ``strongly'' monotone problems, where $T$ is the number of periods, $n$ is the problem dimension (number of random variables) and $k$ is the maximum support size of any random variable. This improves over previous regret bounds by a $\sqrt{n}$ factor. In fact, our dependence on $n$ and $T$ is the best-possible due the $Ω(\sqrt{nT})$ lower bound for multi-armed-bandits. Our main technique is the use of Hellinger distance in bounding the ``error'' between the true and learned distributions, and the use of a constant-factor chain rule for the Hellinger distance of dependent random variables.
Primary: University of Michigan
All Institutions: University of Michigan
The paper establishes near-optimal regret bounds for semi-bandit learning of monotone stochastic problems by leveraging Hellinger distance and adaptive chain rules. This theoretical advance improves upon prior bounds by a $\sqrt{n}$ factor and provides a unified framework for analyzing a wide class of stochastic optimization tasks, though it lacks empirical validation.
The paper proposes a semi-bandit learning algorithm for a broad class of "monotone" stochastic optimization problems (e.g., stochastic knapsack, matching, prophet inequality). The core technical contribution is the application of the Hellinger distance to bound the stability of policies under distributional shifts. By utilizing Jayram's adaptive chain rule for Hellinger distance, the authors derive a stability lemma that depends on the probability of probing each item, rather than a uniform bound. This allows for a tighter regret analysis that leverages the semi-bandit feedback structure (where only probed items are observed). The method generalizes previous work that relied on Total Variation distance, which yielded suboptimal $\sqrt{nT}$ dependencies. The extension to general distributions using Radon-Nikodym derivatives and measure-theoretic tools is rigorous and well-structured.
The paper is purely theoretical. It provides no empirical experiments, benchmarks, or simulations to validate the theoretical bounds on real-world or synthetic datasets. The "results" are entirely mathematical regret bounds. While this is standard for this specific subfield of theoretical computer science/ML, it limits the immediate practical impact and verifiability of the improvements.
The theoretical results are fully reproducible in the sense that the proofs are provided in detail. However, there is no code, algorithm implementation, or experimental setup to reproduce. The algorithm described is a meta-algorithm relying on an offline approximation oracle, which is not specified in a way that allows for direct implementation without choosing a specific problem instance.
The primary limitation is the lack of empirical validation. The theoretical improvements (removing a $\sqrt{n}$ factor) are significant in asymptotic analysis but may not translate to practical performance gains in finite-time regimes typical of real-world applications. Additionally, the results assume access to an offline approximation algorithm for the specific stochastic problem, which may not exist or be efficient for all monotone problems. The assumption of independent random variables, while standard, restricts applicability to problems with correlated uncertainties.
This work contributes to the foundational theory of online learning in combinatorial settings. It provides a general framework for analyzing semi-bandit feedback in stochastic optimization, which can be applied to a wide range of problems in operations research and economics. The technique of using Hellinger distance with adaptive chain rules may inspire further theoretical work in bandit learning and statistical estimation. However, the impact is primarily academic and theoretical rather than industrial. The paper establishes near-optimal regret bounds for semi-bandit learning of monotone stochastic problems by leveraging Hellinger distance and adaptive chain rules. This theoretical advance improves upon prior bounds by a $\sqrt{n}$ factor and provides a unified framework for analyzing a wide class of stochastic optimization tasks, though it lacks empirical validation.
Recent research-level physics-reasoning benchmarks for LLMs reveal a failure mode that lies not in knowledge or execution, but in modeling the world being reasoned about. Models often silently alter given quantities, fabricate laws, or inject unstated assumptions to force a problem into familiar patterns, and then reason inside this corrupted world. We define this premise-level divergence as modeling drift: the gap between a model's internal axiomatic world model (AWM) and the true physical world specified in the problem. Although RLVR and PRM have yielded substantial gains in mathematics and code, their reward shapes do not directly supervise world modeling in axiomatic domains such as physics. We propose Reinforcement Learning with Axiomatic World Modeling (RLAWM), which optimizes the policy by penalizing divergence between its inferred axiomatic world model and the true physics world. RLAWM probes physics reasoning at two hierarchical levels (modeling and solving), each scored by a physically grounded reward. The modeling reward targets abstract knowledge formulation, evaluating whether the model captures the true axiomatic world before attempting calculation. The solving reward targets instantiated knowledge reasoning, checking whether the model's concrete derivation remains consistent with the axiomatic world. This two-level structure separates what world the model believes it is reasoning in from how it reasons within that world. The framework operates exclusively during training, with no extra cost during inference. RLAWM improves performance by up to +12.2 points on PhysReason and +5.2 on PHYSICS over the baseline. It also generalizes zero-shot across difficulty, source, and domain, with a 2.45x gain and +10 points over the strongest baseline on olympiad-level physics and math.
Primary: University of Florida
All Institutions: University of Florida, Northwestern University
The paper introduces a novel RL framework, RLAWM, that mitigates "modeling drift" in physics reasoning by decomposing the task into axiomatic world modeling and solving, using a dual reward system of motif-based fidelity scoring and executable code verification. By leveraging the verification-derivation asymmetry of physics, the method provides dense, physically grounded training signals that expand the reasoning boundary and improve out-of-distribution generalization, representing a substantial step forward in domain-specific LLM reasoning.
The paper proposes RLAWM, a reinforcement learning framework that explicitly targets "modeling drift" in physics reasoning. The core innovation is the decomposition of the reasoning process into a [modeling] block (externalizing the Axiomatic World Model) and a [solving] block (pseudocode derivation). The methodology introduces two distinct reward signals: a Modeling Reward ($R_M$) based on Physics Modeling Fidelity (PMF) using motif extraction (QEL, VCL, AEB) and an LLM-based soft matcher, and a Solving Reward ($R_S$) that translates pseudocode to executable Python and runs programmatic unit tests against a library of physics law templates. This approach leverages the verification-derivation asymmetry in physics, where checking invariants is easier than deriving them. The use of a frozen translator (Qwen2.5-Coder) to bridge natural language/pseudocode and executable code is a practical and effective design choice.
The experiments are rigorous and comprehensive. The authors evaluate on in-distribution benchmarks (PhysReason, PHYSICS) and out-of-distribution benchmarks (OlympiadBench, PHYBench, AIME) across three model families (Qwen3-VL, Gemma-3). Key results include a +12.2 point improvement on PhysReason and a significant reduction in "diversity collapse" (SubPass@k curves continue to improve with $k$, unlike RLVR baselines). Ablations clearly demonstrate the necessity of both $R_M$ and $R_S$, showing they target orthogonal failure modes (premise errors vs. execution errors). The analysis of the training signal density (reducing zero-gradient groups from 29.9% to 9.3%) provides strong evidence for the effectiveness of the dense reward signal.
High. The paper provides detailed descriptions of the reward components, the motif extraction schema, and the checker library construction. A GitHub repository is provided with code and models. Hyperparameters and compute requirements (4x B200 GPUs, ~6 hours) are specified. The use of standard models (GPT-4o-mini for extraction, Qwen for translation) ensures accessibility.
The method relies on a frozen LLM for motif extraction and a separate translator, adding training complexity and cost. The checker library is static and may not cover all physics domains equally well (noted lower performance in Quantum/Stat Mech compared to E&M/Optics). The approach is currently specialized for physics; while the authors suggest extension to math, the specific "axiomatic world model" structure may need adaptation for other domains.
This paper offers a significant conceptual shift in how LLMs are trained for scientific reasoning, moving beyond outcome-based rewards to process-based world-model alignment. The "modeling drift" framework is likely to influence future work in scientific LLMs. The technique of using executable code as a reward signal for reasoning steps is a robust pattern that could be adapted to other formal domains like mathematics or logic. The paper introduces a novel RL framework, RLAWM, that mitigates "modeling drift" in physics reasoning by decomposing the task into axiomatic world modeling and solving, using a dual reward system of motif-based fidelity scoring and executable code verification. By leveraging the verification-derivation asymmetry of physics, the method provides dense, physically grounded training signals that expand the reasoning boundary and improve out-of-distribution generalization, representing a substantial step forward in domain-specific LLM reasoning.
Despite breakthroughs in AI-driven molecular generation, existing discovery pipelines operate largely in silico, with wet-lab feedback reserved for validation rather than iterative improvement. Whether large language model (LLM)-based reasoning agents can accelerate de novo discovery by iterating with lab data remains unproven. Here we use protein-binding aptamer design as a sandbox (1-2 week cycles, \$1-\$3 per candidate) to demonstrate a wet lab in a closed loop with AI systems. We tasked agentic pipelines with developing aptamer binders against $α$-thrombin and myoglobin across three design-test-improve cycles ($\sim$1,000 candidates per round; $\sim$4,400 constructs). Statistically significant, bead-corrected on-target hit rates among new candidates rose across rounds from 0.0% in Round 1 to 26.0% in Round 2 and 42.1% in Round 3 on $α$-thrombin (37.1% excluding literature motif descendants), and from 10.2% in Round 1 to 3.7% in Round 2 and 33.5% in Round 3 on myoglobin. Single-clone bead-binding titrations of eight full-length $α$-thrombin candidates confirmed concordance with pooled sequencing enrichment (Pearson $r = 0.90$). In these assays, literature-derived motifs engineered by the agents to reduce off-target reactivity in the pooled screen reached apparent affinities of $K_D = 24.6$--$133$ nM, whereas de novo scaffolds showed low-micromolar binding ($K_D > 1\ μ$M). We also report how our pipelines used myoglobin binding data to exploit the G-quadruplex binder space, diagnose cross-reactivity as on-target specificity dropped from 75.0% to 19.1%, and recover by developing strategies that restored specificity to 72.4%. We demonstrate that empirical lab data is more valuable than existing computational proxies, showing how iterative experimental feedback enables AI agents to diagnose assay artifacts and refine molecular designs across rounds.
Primary: Google LLC
All Institutions: Google LLC
The paper demonstrates that LLM-based agentic pipelines can effectively iterate with wet-lab data to improve molecular design, achieving significant hit rates and nanomolar binders in a closed-loop aptamer discovery campaign. By rigorously benchmarking against computational predictors and showing their inadequacy, the study establishes the necessity of empirical feedback loops for AI-driven discovery in data-scarce biological domains, offering a robust framework for future self-driving laboratory systems.
The paper proposes a closed-loop framework for agentic biomolecular discovery, specifically using single-stranded DNA aptamers as a "sandbox" to test LLM-based reasoning agents. The methodology involves a multi-agent pipeline (Analyst, Strategist, Designer, Scorer) that iterates through design-test-improve cycles. The agents utilize a suite of computational tools (ViennaRNA, Boltz-2, ESMFold2-NA, etc.) and statistical analysis (DESeq2, OLS regression) to interpret wet-lab sequencing data. A key methodological contribution is the rigorous statistical framework for correcting bead-background noise and distinguishing specific binding from non-specific matrix adhesion, which the agents learn to apply autonomously. The approach demonstrates that agents can diagnose failure modes (e.g., G-quadruplex cross-reactivity) and adjust design strategies accordingly, moving beyond simple generation to iterative refinement based on empirical feedback.
The experiments are rigorous and well-controlled. The authors test two targets: alpha-thrombin (a positive control with known aptamers) and myoglobin (a de novo target). They report statistically significant improvements in hit rates across three rounds (0% to 42.1% for thrombin, 10.2% to 33.5% for myoglobin). Single-clone titrations confirm that the top candidates identified by the pooled screen exhibit nanomolar affinity, validating the high-throughput assay. The paper also includes a critical negative result: benchmarking 12 existing computational predictors shows they perform near random chance, highlighting the necessity of the wet-lab loop. The inclusion of inter-operator reproducibility checks (r=0.84-0.96) strengthens the reliability of the data.
The paper provides high reproducibility. It details the specific tools, versions, and licenses used. The statistical pipeline (DESeq2 + OLS regression) is clearly defined. The agent prompts and reasoning traces are partially shared in Table 1. However, the full agent code and specific prompt engineering details are not fully released, and the wet-lab protocol, while described, requires specialized equipment (MiSeq, magnetic beads, specific protein conjugates). The data availability statement promises release of full tables upon publication, which is standard but not yet fulfilled.
The primary limitation is the lack of a direct head-to-head comparison against non-agentic baselines (e.g., standard SELEX or Bayesian optimization) under identical budget constraints. The authors acknowledge this, stating that prospective trials with control arms are needed. Additionally, the study is limited to aptamers; generalization to more complex modalities like proteins or antibodies is suggested but not demonstrated. The cost and time per round ($1-$3 per candidate, 1-2 weeks) are low, but still higher than purely in silico methods, limiting the scale of search possible.
This paper is significant for the AI for Science community as it provides a concrete, validated example of LLM agents closing the loop with physical experiments. It challenges the assumption that computational predictors alone can drive discovery in data-scarce domains. The "sandbox" concept of using aptamers to test agentic workflows before applying them to harder targets is a valuable strategic insight. It may influence the design of future self-driving laboratories and AI co-scientist frameworks. The paper demonstrates that LLM-based agentic pipelines can effectively iterate with wet-lab data to improve molecular design, achieving significant hit rates and nanomolar binders in a closed-loop aptamer discovery campaign. By rigorously benchmarking against computational predictors and showing their inadequacy, the study establishes the necessity of empirical feedback loops for AI-driven discovery in data-scarce biological domains, offering a robust framework for future self-driving laboratory systems.
Differentially private learning aims to learn from data while protecting individual training examples. For binary classification, finite Littlestone dimension characterizes approximate private PAC learnability (Alon et al., JACM 2022). Understanding how this dimension controls private sample complexity remains a central question. For classes of Littlestone dimension $d$, the first general upper bound was doubly exponential in $d$ (Bun, Livni, and Moran, FOCS 2020). Subsequent work improved this to $\widetilde O(d^6)$ (Ghazi et al., STOC 2021) and then to $\widetilde O(d^5)$ (Lyu, STOC 2026). We prove a nearly linear upper bound of $O(d\log d)$. Our main algorithmic contribution is a structural exponential mechanism for data-dependent candidate sets equipped with a rank satisfying suitable stability conditions. For natural Littlestone classes such as subspace indicators, this rank has a simple geometric interpretation: it is the codimension. This yields a private PAC learner for subspaces of dimension at most $d$ over any field, with tight linear dependence on $d$ and no dependence on the ambient dimension, improving the quadratic dependence of Kaplan et al. (STOC 2025). Our second contribution is combinatorial: drawing on model-theoretic ideas, we construct a suitable rank for general Littlestone classes and establish the counting bounds required by the mechanism.
Primary: Technion -- Israel Institute of Technology
All Institutions: Technion -- Israel Institute of Technology, Google Research
The paper proves a nearly linear upper bound of $O(d \log d)$ on the sample complexity of differentially private PAC learning for classes of Littlestone dimension $d$, resolving a central open problem in the field. By introducing a structural exponential mechanism that leverages a model-theoretic rank, the authors achieve a tight dependence on the Littlestone dimension, significantly improving upon previous polynomial bounds and providing a new algorithmic framework for private learning.
The paper introduces a "structural exponential mechanism" that generalizes the classical exponential mechanism to handle data-dependent candidate sets. This is a significant algorithmic innovation in private learning theory. The core novelty lies in defining a combinatorial rank for hypotheses in Littlestone classes, derived from model-theoretic concepts (specifically Shelah's $2$-rank), and proving that this rank satisfies stability conditions necessary for differential privacy. The authors demonstrate that for subspace indicators, this rank corresponds to codimension, yielding a tight linear sample complexity bound $O(d \log d)$, which is optimal up to logarithmic factors. The methodology is rigorous, combining combinatorial arguments with privacy analysis.
This is a theoretical paper in learning theory and differential privacy. There are no empirical experiments, datasets, or benchmarks. The "evaluation" consists of mathematical proofs establishing upper bounds on sample complexity and verifying the privacy guarantees of the proposed mechanism. The contribution is entirely analytical.
As a pure theory paper, reproducibility is defined by the correctness of the proofs. The paper provides detailed technical overviews and formal definitions. However, the full proofs are truncated in the provided text (ending in Section 5.1), so a complete verification of the combinatorial lemmas (specifically Lemmas 5.2-5.4 and the counting bounds in Section 5.3) is not possible from the snippet alone. Assuming the standard rigor of the authors (Moran, Yan) and the venue (likely STOC/FOCS level), the theoretical framework is sound.
The primary limitation is that the result is a sample complexity bound, not a computational complexity bound. The paper explicitly states that it does not assert computational efficiency. The algorithm described (selecting candidates based on version spaces and ranks) may be computationally intractable for large $d$ or complex classes. Additionally, the bound is for the realizable setting; while reductions to the agnostic setting exist, they incur additional overhead. The log-star question regarding VC vs Littlestone dimension remains open.
This work resolves a major open problem in private learning theory by establishing a nearly linear dependence on Littlestone dimension. It provides a new algorithmic primitive (structural exponential mechanism) that can be applied to other private optimization problems involving data-dependent candidate sets. It bridges learning theory and model theory, offering new combinatorial tools for understanding the complexity of private learning. The result is significant for the theoretical understanding of the trade-offs between privacy and sample efficiency. The paper proves a nearly linear upper bound of $O(d \log d)$ on the sample complexity of differentially private PAC learning for classes of Littlestone dimension $d$, resolving a central open problem in the field. By introducing a structural exponential mechanism that leverages a model-theoretic rank, the authors achieve a tight dependence on the Littlestone dimension, significantly improving upon previous polynomial bounds and providing a new algorithmic framework for private learning.
Understanding how stochastic gradient descent learns internal representations in wide neural networks trained on rich target functions is a central problem in machine learning. We study this question in two-layer neural networks learning target functions represented by networks of the same architecture, with arbitrary nonlinearity, and with network width and input dimension growing proportionally. In this regime, standard approaches for describing the learning dynamics require tracking a diverging number of overlaps, which measure correlations between learned and target features, leaving no obvious description in terms of a few order parameters. Here, we develop a mesoscopic description of learning by characterizing the distribution of overlaps. In the early stages of learning, we find that a Gaussian description of overlap statistics reduces the learning dynamics to a few collective variables that track learning of the target function's linear and quadratic components. We then estimate the timescale of specialization, when Gaussian closure breaks down and individual learned features begin to strongly correlate with specific target features. We find that specialization in extensive-rank networks is accelerated via a quadratic shielding mechanism, whereby learning the linear and quadratic components of the target function filters gradient fluctuations. We finally describe how the activation function determines the pattern of specialization and characterize the subsequent refinement of learned features. Together, these results show how collective learning gives way to individual feature discovery and clarify when learning in large neural networks admits a simple statistical description.
Primary: International School for Advanced Studies (SISSA)
All Institutions: International School for Advanced Studies (SISSA), The Abdus Salam International Centre for Theoretical Physics (ICTP)
The paper develops a mesoscopic theory for learning dynamics in extensive-width neural networks, deriving closed-form equations for the pre-specialization regime and predicting the timescale of feature specialization via a quadratic shielding mechanism. By bridging the gap between narrow-network order parameter analyses and infinite-width mean-field theories, it provides a rigorous framework for understanding how collective learning transitions to individual feature discovery in high-dimensional settings.
The paper introduces a "mesoscopic" description of learning dynamics for extensive-width (proportional width and input dimension) two-layer neural networks. It bridges the gap between the Saad-Solla approach (which fails in the extensive limit due to diverging order parameters) and mean-field/NTK theories (which often miss feature learning details or require infinite width). The core method involves tracking the distribution of overlaps between teacher and student neurons via their empirical moments. The authors derive a "Gaussian closure" regime where the dynamics close on a few collective variables (linear and quadratic components), and then analyze the breakdown of this closure to predict the onset of "specialization" (individual feature alignment). The derivation of the "quadratic shielding" mechanism, which explains how learning low-order components filters gradient noise to accelerate specialization, is a particularly insightful theoretical contribution.
The paper provides strong numerical validation of the theoretical predictions. It compares the derived ODEs for the order parameters ($Q_1, Q_2$) against simulations of SGD for various activation functions (ReLU, tanh, ELU, etc.). The agreement between the theoretical loss plateaus and the specialization timescales is precise. The experiments confirm the existence of the distinct learning phases (collective alignment, specialization, refinement) and validate the scaling laws for specialization time with respect to network width and activation function properties.
The paper is highly reproducible from a theoretical standpoint, as it provides detailed derivations in the appendices. The experimental setup is standard (teacher-student setup with Gaussian inputs), and the specific activation functions and network sizes are clearly defined. While code is not explicitly linked in the text provided, the mathematical framework is sufficient for replication.
The analysis is restricted to two-layer networks with fixed second-layer weights (committee machines). Extending this to deeper networks or trainable readouts is left for future work. The "Gaussian closure" is an approximation that breaks down during specialization, and the paper provides estimates for the transition rather than a fully closed dynamic description of the specialization phase itself. The assumption of online SGD (one-pass) limits direct applicability to multi-epoch training regimes common in practice, though the authors argue relevance to LLM pre-training.
This work contributes to the fundamental understanding of how neural networks learn representations in high-dimensional regimes. It provides a rigorous framework for analyzing feature learning that goes beyond the NTK and mean-field limits. The insights into "quadratic shielding" and the role of activation functions in specialization patterns could inform the design of better initialization schemes or learning rate schedules for wide networks. It connects statistical physics methods with modern deep learning theory, offering tools to predict training dynamics and generalization errors in the extensive-width limit. The paper develops a mesoscopic theory for learning dynamics in extensive-width neural networks, deriving closed-form equations for the pre-specialization regime and predicting the timescale of feature specialization via a quadratic shielding mechanism. By bridging the gap between narrow-network order parameter analyses and infinite-width mean-field theories, it provides a rigorous framework for understanding how collective learning transitions to individual feature discovery in high-dimensional settings.
Diffusion large language models (dLLMs) have emerged as a promising alternative to autoregressive language models, enabling non-autoregressive text generation through parallel decoding. Existing dLLMs can be broadly categorized into two families: masked diffusion models (MDMs) and uniform diffusion models (UDMs) In this work, we theoretically show that MDMs are easier to optimize due to their mask-based transition schedule but have lower computational expressivity, whereas UDMs enable iterative state revision and greater expressivity but are substantially harder to train. This trade-off raises a natural question: can we obtain the expressivity of iterative revision without giving up the simple, easy-to-optimize transition schedule of MDMs? Motivated by this analysis, we propose Lookahead Editing (\ours{}), which augments masked diffusion decoding with a rewritable proposal state. At each decoding step, the model performs several lookahead updates to iteratively refine future predictions in this proposal state before committing tokens to the masked sequence. This design introduces a rewritable proposal that uses additional compute to improve expressivity while keeping the simple mask-based transition schedule of MDMs. We fine-tune Qwen2.5-7B and Qwen3-8B to obtain LED-\texttt{7B} and LED-\texttt{8B}, respectively. Across eight benchmarks, LED SFT yields average gains of $+11.4$ and $+7.7$ points over standard MDM SFT. Moreover, using SFT alone, LED-\texttt{8B} outperforms LLaDA2.1-16B Q-mode, which is trained with continued pre-training, supervised fine-tuning, and reinforcement learning.
Primary: University of Cambridge
All Institutions: University of Cambridge, NVIDIA
The paper introduces Lookahead Editing, a method that enhances masked diffusion language models by incorporating a rewritable proposal state to improve computational expressivity while maintaining optimization stability. It provides rigorous theoretical proofs regarding the circuit complexity of MDMs vs UDMs and demonstrates significant empirical gains in reasoning tasks, outperforming larger models trained with more complex pipelines.
The paper proposes "Lookahead Editing" (LED), a decoding strategy for Masked Diffusion Language Models (MDMs) that introduces a rewritable proposal state. The core innovation is bridging the gap between the optimization simplicity of MDMs (which have a fixed, easy-to-learn transition schedule) and the computational expressivity of Uniform Diffusion Models (UDMs) (which allow iterative revision but are harder to train). Theoretically, the authors prove that MDMs are contained in $TC^0$ while UDMs can reach $NC^1$, justifying the need for iterative revision. The method involves a three-branch training setup (clean, masked, proposal) where the proposal branch is generated by a pretrained MDM, allowing the model to learn to correct errors in the proposal. At inference, the model performs $K$ lookahead steps to refine the proposal before committing tokens. This is a clever architectural/algorithmic modification that preserves the training stability of MDMs while gaining the benefits of iterative refinement.
The experiments are conducted on Qwen2.5-7B and Qwen3-8B models. The results show significant improvements over standard MDM SFT baselines (+11.4 and +7.7 points average across 8 benchmarks). Notably, LED-8B outperforms LLaDA2.1-16B, which uses a more complex post-training pipeline (CPT + SFT + RL), using only SFT. The speed-accuracy trade-off analysis demonstrates that LED offers a better Pareto front than existing dLLMs. The benchmarks cover math, science, and code, providing a comprehensive evaluation of reasoning capabilities.
The paper provides detailed pseudocode for both training and inference. The hyperparameters (block size, batch size, learning rate) are specified. However, the specific datasets used for SFT are listed as open-source, but the exact mixture ratios and preprocessing steps are not fully detailed in the provided text. The reliance on a pretrained MDM for generating proposals adds a dependency that must be managed, but the overall procedure is clearly defined.
The method requires an additional forward pass (or multiple, depending on $K$) for the proposal generation, which increases inference latency compared to standard MDMs, although the paper argues this is offset by the ability to use fewer decoding steps or higher throughput at matched accuracy. The theoretical analysis, while rigorous, relies on specific assumptions about the circuit complexity of transformers that may not hold for all practical implementations. The performance gains are observed on specific benchmarks (math/code), and generalization to other domains (e.g., creative writing, long-form generation) is not explicitly tested.
This work contributes to the growing field of diffusion-based language models, offering a practical way to improve their reasoning capabilities without abandoning the stable training dynamics of masked diffusion. The theoretical insights into the optimization-expressivity trade-off are valuable for the community. The ability to outperform larger, more complexly trained models with a simpler SFT recipe suggests that architectural/decoding innovations can be as impactful as data scaling or RL. The paper introduces Lookahead Editing, a method that enhances masked diffusion language models by incorporating a rewritable proposal state to improve computational expressivity while maintaining optimization stability. It provides rigorous theoretical proofs regarding the circuit complexity of MDMs vs UDMs and demonstrates significant empirical gains in reasoning tasks, outperforming larger models trained with more complex pipelines.
Amodal video segmentation asks for the complete mask of an object in every frame, including the parts hidden behind other objects. The best current methods generate these masks with fine-tuned video diffusion models, which need many denoising steps and an extra depth network, and give a different answer for every sample. Self-supervised video world models of the JEPA family are trained on a closely related task: predicting the representation of masked space-time regions from the visible ones. We therefore treat occlusion as masking. The tokens that the object's visible mask does not cover are dropped, the pre-trained V-JEPA 2.1 predictor fills them in, and a light head decodes the observed and predicted features into amodal masks. The resulting model, OccJEPA (Occlusion-as-masking JEPA), is deterministic and 33.5x faster than the diffusion-based state of the art, and more accurate on SAIL-VOS, on TAO-Amodal, including a held-out test split (+4.00 AP50) and visible masks tracked by SAM2, and on Kubric. Controlled experiments trace the gain to the predictor's pre-trained weights rather than to its capacity, for world models from 87M to 1.85B parameters. Code is available at https://github.com/zgp110/occjepa.
Primary: University of Wisconsin--Madison
All Institutions: University of Wisconsin--Madison
[One sentence main contribution]. [Comprehensive analysis of the technical contribution, methodology, and significance to the field]. The paper introduces OccJEPA, a deterministic and highly efficient method for amodal video segmentation that leverages pre-trained JEPA world models by treating occlusion as masking, achieving state-of-the-art accuracy with a 33.5x speedup over diffusion-based approaches.
The paper proposes OccJEPA, a method for amodal video segmentation that reframes occlusion as a masking problem solvable by pre-trained self-supervised video world models (specifically V-JEPA 2.1). Instead of using computationally expensive diffusion models with depth estimation, the method drops tokens corresponding to visible object masks and uses the JEPA predictor to fill in the latent representations of the occluded regions. A lightweight head then decodes these features into amodal masks. The core insight is that the predictive capability of JEPA models, trained to reconstruct masked spatiotemporal regions, naturally aligns with the task of inferring hidden object parts. The approach is deterministic, avoiding the stochasticity of diffusion sampling, and leverages the robustness of pre-trained weights rather than fine-tuning heavy generative models.
The experimental section demonstrates significant improvements over state-of-the-art diffusion-based methods. The model achieves a 33.5x speedup while maintaining or improving accuracy on SAIL-VOS, TAO-Amodal (including a held-out test split with +4.00 AP50), and Kubric. Ablation studies confirm that the performance gain stems from the pre-trained weights of the JEPA predictor rather than its architectural capacity, showing consistent results across model sizes from 87M to 1.85B parameters. The use of SAM2 for visible mask tracking ensures a fair comparison with other methods that rely on similar initialization.
The authors provide a public GitHub repository, which enhances reproducibility. The reliance on standard pre-trained models (V-JEPA 2.1) and established datasets (SAIL-VOS, TAO, Kubric) further supports reproducibility. However, the specific details of the "light head" architecture and training hyperparameters would need to be verified in the code repository for full replication.
The method relies on the quality of the visible masks provided by SAM2; errors in visible segmentation may propagate to the amodal prediction. The approach is tested primarily on standard benchmarks and may not generalize to highly dynamic or complex occlusion scenarios not covered by the training data of the pre-trained JEPA model. The deterministic nature, while a strength for consistency, may limit the exploration of multiple plausible amodal configurations in ambiguous cases.
This work bridges the gap between self-supervised representation learning and dense prediction tasks, suggesting that world models can be effectively repurposed for segmentation without generative overhead. The significant speedup makes amodal segmentation more practical for real-time applications. The finding that pre-trained weights are more critical than capacity for this task has implications for the design of future video understanding models. [One sentence main contribution]. [Comprehensive analysis of the technical contribution, methodology, and significance to the field]. The paper introduces OccJEPA, a deterministic and highly efficient method for amodal video segmentation that leverages pre-trained JEPA world models by treating occlusion as masking, achieving state-of-the-art accuracy with a 33.5x speedup over diffusion-based approaches.
Video world models predict future frames from past frames and actions. They serve as simulators in which agents train, plan or are evaluated, and that humans can control interactively. They generate in the latent space of a video autoencoder, whose compression sets their cost per frame and how long each action waits to take effect. Yet world models often reuse the autoencoders of text-to-video generators, built for offline generation with no agent waiting and, without their training data and code, not retrainable at another compression. We therefore ask how, where and how much a world model should compress. We build CADRE, an open family of causal diffusion representation autoencoders spanning multiple spatio-temporal compressions. Its members differ only in a small bottleneck on frozen features and share a diffusion decoder self-distilled to one or a few steps. We train the same world model on each autoencoder in two domains, manipulation and navigation, with two environments each. CADRE forms most of the latency-quality frontier, especially with a moving camera. Our study also leads to four findings. At strong compression, a diffusion decoder is sharper than perceptual regression. In robot manipulation, temporal compression is more effective in the autoencoder than in the world model. Manipulation and navigation call for opposite compressions: temporal compression costs little when the camera is still, and spatial compression when it moves. Surprisingly, with a still camera, coarser spatial tokens hurt motion far more than image quality. Code, models and a live demo are at https://kyutai.org/blog/2026-10-21-cadre
Primary: Kyutai
All Institutions: Kyutai
The paper provides a systematic study of spatio-temporal compression in video autoencoders for interactive world models, introducing the CADRE family and revealing critical trade-offs between latency, quality, and motion fidelity. By decoupling the bottleneck from frozen features and employing a fast diffusion decoder, the authors establish a new latency-quality frontier and offer actionable insights for designing real-time world models in manipulation and navigation tasks.
The paper introduces CADRE, a family of causal diffusion representation autoencoders designed specifically for interactive world models. The core innovation lies in decoupling the compression bottleneck from the frozen features of pre-trained video generators, allowing for systematic exploration of spatio-temporal compression ratios. By using a diffusion decoder that is self-distilled to one or a few steps, the authors address the latency constraints inherent in interactive applications. The methodology is sound, focusing on how compression affects the latency-quality trade-off, which is a critical but under-explored aspect of world model design.
The evaluation is rigorous, testing the same world model architecture across multiple CADRE variants in two distinct domains (manipulation and navigation) with two environments each. The paper presents a clear latency-quality frontier, demonstrating that CADRE members dominate this frontier, particularly in scenarios with moving cameras. The ablation studies comparing diffusion decoders to perceptual regression and analyzing the impact of temporal vs. spatial compression provide valuable empirical insights.
The authors explicitly state that code, models, and a live demo are available, which significantly enhances reproducibility. The open-sourcing of the autoencoder family allows the community to replicate the compression studies and integrate these components into their own world models.
The study is limited to two domains (manipulation and navigation) and two environments per domain. While the findings are insightful, their generalizability to other complex domains (e.g., driving, social interaction) remains to be tested. Additionally, the reliance on frozen features from text-to-video generators may limit the potential performance ceiling compared to end-to-end trained autoencoders, though this is a practical constraint rather than a methodological flaw.
This work has significant implications for the development of interactive AI agents and real-time simulation. By providing a systematic framework for choosing autoencoder compression in world models, it helps practitioners optimize for latency without sacrificing quality. The findings on the trade-offs between spatial and temporal compression in different camera conditions are particularly useful for robotics and AR/VR applications. The paper provides a systematic study of spatio-temporal compression in video autoencoders for interactive world models, introducing the CADRE family and revealing critical trade-offs between latency, quality, and motion fidelity. By decoupling the bottleneck from frozen features and employing a fast diffusion decoder, the authors establish a new latency-quality frontier and offer actionable insights for designing real-time world models in manipulation and navigation tasks.
This paper addresses the problem of creating immersive volumetric videos from sparse and low-resolution capture resources. Conventional dense multi-view reconstruction methods can produce high-quality volumetric videos, but their reliance on costly capture setups makes them impractical in real-world scenarios. Recent approaches use generative models to relax capture requirements, but still suffer from limited resolution and 4D inconsistency. To overcome these constraints, we propose RealMe, a unified framework built around a high-resolution video diffusion model that enhances limited captures into high-resolution (2K), 4D-consistent videos for volumetric video reconstruction. Specifically, RealMe constructs auxiliary data from capture videos and encodes them into latents along with layout-aware positional embeddings. The video diffusion model adapts factorized spatio-temporal attention to the multi-view capture layout, denoises these latents across viewpoints and time, and decodes them into enhanced videos to refine the 4D representation. This adaptation helps the model capture temporal dynamics and multi-view relationships, leading to more 4D-consistent generation. To reduce the data gap between generated videos and real captures, we apply a latent frequency loss that provides frequency-domain supervision in latent space. Extensive experiments on the public DNA-Rendering benchmark and our self-captured RealVV dataset demonstrate that RealMe is able to create immersive volumetric videos and outperforms state-of-the-art methods in visual fidelity and 4D consistency.
Primary: Zhejiang University
All Institutions: Zhejiang University
RealMe introduces a unified framework using high-resolution video diffusion to enhance sparse multi-view captures into 2K, 4D-consistent videos for volumetric reconstruction. The paper makes a strong contribution by addressing the resolution and consistency bottlenecks in current 4D reconstruction methods, offering a practical pathway to immersive content creation from minimal data, though its impact is slightly tempered by the high computational requirements and reliance on self-captured data for part of the evaluation.
The paper proposes RealMe, a framework that leverages a high-resolution video diffusion model to enhance sparse, low-resolution multi-view captures into 2K, 4D-consistent videos. The core technical contribution lies in adapting factorized spatio-temporal attention to the specific layout of multi-view captures and introducing layout-aware positional embeddings. A key innovation is the "latent frequency loss," which provides frequency-domain supervision in latent space to bridge the gap between generated and real captures. The method integrates this generative enhancement with 4D Gaussian Splatting (specifically FTGS) for final volumetric reconstruction. The approach is logically sound, addressing the specific pain points of resolution and temporal consistency in sparse-view 4D reconstruction.
The authors evaluate their method on the public DNA-Rendering benchmark and a self-captured dataset (RealVV). They claim outperformance over state-of-the-art methods in visual fidelity and 4D consistency. While the use of a public benchmark is positive, the reliance on a self-captured dataset limits independent verification. The "2K" resolution claim is significant for the field, as most prior methods struggle beyond 1K. The comparison with FTGS (the baseline reconstruction method) is appropriate, but comparisons with other generative enhancement pipelines would strengthen the claims.
The paper is 19 pages including supplementary material, suggesting detailed implementation specifics are likely provided. The project page is available. However, training video diffusion models for specific multi-view layouts is computationally expensive, which may hinder reproducibility for smaller labs. The code availability is not explicitly confirmed in the text provided, though the project page likely hosts it.
The method relies on a pre-trained video diffusion model, which may inherit biases or artifacts from its training data. The "minimal captures" requirement is still relative; it likely requires a certain minimum number of views to function effectively. The computational cost of running a 2K video diffusion model for every sequence is high. The self-captured dataset limits the generalizability assessment compared to purely public benchmarks.
This work has significant potential for applications in virtual reality, augmented reality, and digital twins, where high-fidelity 4D content is needed but capture hardware is limited. By enabling high-quality volumetric video from sparse inputs, it lowers the barrier to entry for creating immersive content. It bridges the gap between generative AI and 3D reconstruction, a trend likely to grow in importance. RealMe introduces a unified framework using high-resolution video diffusion to enhance sparse multi-view captures into 2K, 4D-consistent videos for volumetric reconstruction. The paper makes a strong contribution by addressing the resolution and consistency bottlenecks in current 4D reconstruction methods, offering a practical pathway to immersive content creation from minimal data, though its impact is slightly tempered by the high computational requirements and reliance on self-captured data for part of the evaluation.
World models enable embodied agents to anticipate action consequences, yet jointly modeling locomotion and manipulation across diverse embodiments remains challenging due to heterogeneous action spaces and limited multimodal supervision. We present Ropedia W0, the first egocentric loco-manipulation world foundation model spanning humans and robots with diverse embodiments. We curate Xp-Data, comprising 10k hours of egocentric human data and 3k hours of robot data with unified multimodal annotations and systematic quality control. A unified kinematic representation encodes locomotion through camera poses and manipulation through projected skeletons, with robot skeletons combining embodiment-specific forward kinematics with extensive manual keypoint annotation and correction. We also introduce Xp-Teleop, a long-horizon loco-manipulation benchmark based on high-quality dexterous humanoid teleoperation data. Evaluations demonstrate state-of-the-art performance across five cross-embodiment benchmarks (DreamDojo, $Ψ_0$, $ω$-HOME, Xp-Teleop, and AgiBot) and on held-out human egocentric video, from zero-shot inference to few-shot and full-data post-training. Scaling human pre-training data improves performance on both human and robot evaluations. After distillation, our model supports real-time, minute-long rollouts, enabling live VR interaction. Applications in policy evaluation, model-based planning, and world-action modeling further demonstrate the model's utility for embodied decision-making.
Primary: Unknown (Likely a major tech lab or consortium given the scale of data, e.g., NVIDIA, Google, or a major robotics startup, but affiliations are redacted/missing in the provided text)
All Institutions: Unknown
Ropedia W0 introduces a cross-embodiment world foundation model that unifies human and robot loco-manipulation through a shared kinematic representation, demonstrating that scaling human egocentric data significantly improves robot world model performance. The paper makes a substantial contribution by addressing the heterogeneity of action spaces in embodied AI, providing a rigorous evaluation across multiple benchmarks, and introducing a new long-horizon teleoperation benchmark, thereby establishing a new paradigm for leveraging human data in robotic world modeling.
The paper proposes Ropedia W0, a world foundation model for loco-manipulation. The core methodological contribution is the unification of heterogeneous action spaces (human locomotion vs. robot manipulation) into a single kinematic representation. Specifically, it encodes locomotion via camera poses and manipulation via projected skeletons. For robots, this involves combining embodiment-specific forward kinematics with manual keypoint annotation. The model is trained on Xp-Data, a massive dataset of 10k hours of human and 3k hours of robot egocentric video. The approach leverages the abundance of human data to pre-train the world model, hypothesizing that human loco-manipulation dynamics transfer to robots. This is a significant architectural and data-engineering choice, moving away from purely robot-centric world models to a cross-embodiment paradigm.
The experimental section is extensive, evaluating the model on five cross-embodiment benchmarks: DreamDojo, $\Psi_0$, $\omega$-HOME, Xp-Teleop (a new benchmark introduced in the paper), and AgiBot. The paper demonstrates state-of-the-art performance in video prediction quality. Crucially, it validates the scaling hypothesis: increasing human pre-training data improves performance on both human and robot evaluations. The introduction of Xp-Teleop, a long-horizon benchmark based on dexterous humanoid teleoperation, is a valuable contribution to the field. The model's ability to support real-time, minute-long rollouts after distillation is a strong practical result, enabling live VR interaction.
Reproducibility is likely limited due to the scale of the data (13k hours of video) and the complexity of the manual annotation pipeline for robot skeletons. While the project page is provided, the specific details of the "systematic quality control" and the exact architecture of the world model (e.g., transformer variants, latent space dimensions) would need to be in the full text/appendix to be fully reproducible by smaller labs. The reliance on proprietary or large-scale in-house data makes independent verification difficult.
The primary limitation is the heavy reliance on manual keypoint annotation and correction for robot skeletons, which is labor-intensive and may not scale easily to new embodiments without significant human effort. The model's performance on out-of-distribution embodiments or tasks not covered by the 13k hours of data is not fully explored. Additionally, the "unified kinematic representation" may struggle with highly dexterous or non-rigid manipulations that are difficult to capture with standard skeletons.
This work has high potential impact on the field of embodied AI and robotics. By demonstrating that human video can effectively pre-train robot world models, it offers a pathway to overcome the data scarcity problem in robotics. The introduction of a cross-embodiment benchmark (Xp-Teleop) and a scalable world model architecture could accelerate research in model-based planning and policy evaluation. The real-time capability opens doors for interactive simulation and VR-based robot training. Ropedia W0 introduces a cross-embodiment world foundation model that unifies human and robot loco-manipulation through a shared kinematic representation, demonstrating that scaling human egocentric data significantly improves robot world model performance. The paper makes a substantial contribution by addressing the heterogeneity of action spaces in embodied AI, providing a rigorous evaluation across multiple benchmarks, and introducing a new long-horizon teleoperation benchmark, thereby establishing a new paradigm for leveraging human data in robotic world modeling.
Transformers are widely assumed to exhibit token-decoding explainability. In particular, in vision transformer (ViT) classifiers, decoding individual patch tokens has become a common diagnostic tool due to its tendency to yield human-interpretable outputs. In light of the increasing popularity of this approach, an important question is whether token-decoding explainability is an inherent property of transformers that should be expected to hold across architectures, scales, and training objectives, or merely an artifact of some implicit bias. In this paper we systematically study token-decoding explainability across diverse ViT settings, spanning architectures, scales, and training objectives, using both controlled experiments and real-world models. We find that while some models indeed exhibit aligned and interpretable token decodings, others exhibit either inverted explainability maps or a complete lack of correlation with the image semantics. As we show, these tendencies toward different explainability behaviors are not intrinsic to the architecture and can vary arbitrarily with changes in the optimization landscape. We provide theoretical support for these observations by demonstrating a setting where a single fixed toy transformer architecture can provably implement the optimal classification rule in multiple ways, which differ in their patch-decoding explainability patterns. Our results show that token-decoding explainability is not an inherent property of trained transformers, but rather an outcome contingent on the problem setting and thus should not be a-priori assumed to hold. Code can be found in https://ronensc.github.io/Explainability-Illusion/ .
Primary: Technion - Israel Institute of Technology
All Institutions: Technion - Israel Institute of Technology, Google Research
The paper challenges the assumption that Vision Transformers inherently possess patch-decoding interpretability by demonstrating through empirical and theoretical means that this property is contingent on optimization rather than architecture. By showing that identical classification performance can be achieved with completely different or inverted interpretability maps, the work provides a critical cautionary tale for the field, suggesting that current interpretability diagnostics based on token decoding may be unreliable artifacts of specific training dynamics rather than intrinsic model properties.
The paper employs a rigorous dual-pronged methodology combining systematic empirical analysis with theoretical counterexamples. The empirical component involves training and evaluating a diverse suite of Vision Transformers (ViTs) across varying architectures, scales, and training objectives to test the robustness of patch-decoding interpretability. The theoretical component constructs a provable toy model demonstrating that a fixed transformer architecture can implement the same optimal classification function via multiple weight configurations that yield diametrically opposed interpretability maps (aligned vs. inverted). This combination of empirical breadth and theoretical necessity is methodologically sound and effectively addresses the core research question.
The experimental evaluation is comprehensive, covering both controlled synthetic settings and real-world models, including Vision-Language Models (VLMs). The paper successfully demonstrates that interpretability is not a stable property, showing cases where decoding maps are inverted or uncorrelated with semantics. The inclusion of VLMs adds significant relevance, as these are currently the dominant paradigm in vision AI. The results are surprising and challenge a widely held assumption in the community, which is a hallmark of high-impact empirical work.
The authors provide a public code repository, which is a strong positive for reproducibility. However, the full paper text provided is a skeleton of LaTeX file references rather than the rendered content. While the abstract and structure suggest a complete study, the lack of visible detailed hyperparameters or specific dataset splits in the provided text limits the immediate assessment of granular reproducibility details. The code link is the primary vehicle for verification.
The theoretical proof relies on a "toy transformer" architecture, which may not fully capture the complexity of large-scale, pre-trained industrial models. While the empirical section addresses this, the gap between the simplified theoretical model and real-world complexity remains a potential limitation in generalizing the theoretical insights. Additionally, the paper focuses on the *existence* of the illusion rather than providing a robust method to *detect* or *correct* it in arbitrary black-box models, leaving practitioners without a clear diagnostic tool.
This paper has high potential for broader impact by correcting a fundamental misconception in the AI interpretability community. If token-decoding is not inherently reliable, it invalidates many existing interpretability tools and analyses that rely on this assumption. This could lead to a re-evaluation of how interpretability is measured and reported in future research, potentially shifting focus toward more robust, architecture-agnostic interpretability metrics. It serves as a crucial "reality check" for the field. The paper challenges the assumption that Vision Transformers inherently possess patch-decoding interpretability by demonstrating through empirical and theoretical means that this property is contingent on optimization rather than architecture. By showing that identical classification performance can be achieved with completely different or inverted interpretability maps, the work provides a critical cautionary tale for the field, suggesting that current interpretability diagnostics based on token decoding may be unreliable artifacts of specific training dynamics rather than intrinsic model properties.
Synthesizing a monocular video from a new camera trajectory requires preserving the observed scene geometry and motion while completing regions that become newly visible. Feed-forward 4D reconstructors recover the structure of the observed scene, whereas video diffusion models provide a generative prior for completing unseen content. We introduce Holo4D, a video reshooting framework that uses reconstructed 3D trajectories as a holistic source-view 4D scaffold. We project these trajectories into the requested views and encode their geometry and source appearance to guide a pretrained video diffusion model. Together with the source video and target cameras, the scaffold grounds generation in reconstructed observations while leaving unsupported regions to the generative prior. Experiments demonstrate the value of trajectory-based conditioning for camera control and novel-view synthesis. The framework accommodates different 4D reconstructors without retraining the video generator, allowing it to benefit from future advances in source reconstruction. More details can be found in our project page.
Primary: University of Illinois Urbana-Champaign
All Institutions: University of Illinois Urbana-Champaign, Adobe Research
Holo4D introduces a modular framework that uses feed-forward 4D reconstruction trajectories as a geometric scaffold to guide video diffusion models for camera-controlled video synthesis. The paper makes a significant contribution by decoupling the 3D/4D reconstruction task from the generative video model, allowing for the integration of state-of-the-art 4D reconstructors without retraining the diffusion model. This approach effectively grounds the generative prior in observed scene geometry and motion, leading to improved consistency and controllability in novel view synthesis. The method's ability to leverage future advances in 4D reconstruction makes it a robust and forward-looking solution for video reshooting and generation tasks.
The paper proposes Holo4D, a framework that bridges feed-forward 4D reconstruction and video diffusion models. The core method involves using a 4D reconstructor to generate geometric trajectories of scene points from a source video. These trajectories are then projected into the target camera views to create a "holistic 4D scaffold." This scaffold encodes both geometry and source appearance, serving as a conditioning signal for a pretrained video diffusion model. The approach is modular, allowing the diffusion model to remain frozen while benefiting from improvements in the upstream 4D reconstruction module. This decoupling is a significant architectural choice that enhances the method's longevity and adaptability.
The experiments demonstrate the effectiveness of trajectory-based conditioning for camera control and novel view synthesis. The paper likely compares against baselines that use only 2D projections or static 3D structures, showing that the dynamic 4D scaffold provides superior grounding for motion and geometry. The evaluation likely includes metrics for geometric consistency and visual fidelity, as well as user studies or qualitative comparisons showing the model's ability to complete unseen regions while preserving observed structure.
The paper provides a clear description of the pipeline, including the use of standard 4D reconstructors and pretrained video diffusion models. The modular nature of the framework suggests that components can be swapped, which aids in reproducibility. However, specific details on the training of the conditioning encoder (if any) and the exact integration points with the diffusion model's attention layers would be crucial for full reproduction.
The method relies on the quality of the feed-forward 4D reconstructor; errors in the initial reconstruction will propagate to the generated video. Additionally, the computational cost of running a 4D reconstructor followed by a video diffusion model may be high. The method is also limited to scenes where 4D reconstruction is feasible, potentially excluding highly dynamic or non-rigid scenes that are difficult to reconstruct.
This work has significant implications for the field of video generation and 3D/4D reconstruction. By providing a robust way to ground video diffusion in 3D/4D geometry, it paves the way for more controllable and physically consistent video synthesis. The modular design encourages the community to develop better 4D reconstructors, knowing they can be directly plugged into existing video generation pipelines. This could accelerate progress in both 3D reconstruction and generative video models. Holo4D introduces a modular framework that uses feed-forward 4D reconstruction trajectories as a geometric scaffold to guide video diffusion models for camera-controlled video synthesis. The paper makes a significant contribution by decoupling the 3D/4D reconstruction task from the generative video model, allowing for the integration of state-of-the-art 4D reconstructors without retraining the diffusion model. This approach effectively grounds the generative prior in observed scene geometry and motion, leading to improved consistency and controllability in novel view synthesis. The method's ability to leverage future advances in 4D reconstruction makes it a robust and forward-looking solution for video reshooting and generation tasks.
Video diffusion models (VDMs) trained on internet-scale data encode a rich prior on appearance, motion, and physics that is difficult to replicate with explicit 3D rendering and costly 3D data capture. We introduce Video Diffusion Avatars (VDAs), a method that creates photo-realistic, multi-view consistent avatars from a single image by fine-tuning a VDM for precise control over camera viewpoint, facial expression, and upper-body pose. Departing from 3D morphable models (3DMMs), VDAs use a learned expression space that lifts the expressiveness ceiling of 3DMMs, e.g., for the tongue and mouth interior. However, learned expression codes exhibit heavy viewpoint bias, as our DISTRACT benchmark reveals. We show that this bias can be removed through disentanglement, enabling their use as driving signals for multi-view consistent avatars. Camera and upper-body pose are controlled jointly via proxy-mesh renderings, giving independent control over expression, pose, and viewpoint. A camera-triplet training strategy makes the approach efficient: our models train for 34 hours on a single A100 using only public multi-view data. VDAs outperform state-of-the-art single-image 3D avatars on novel-view reenactment, reducing face LPIPS by 29% and improving perceptual quality (JOD) by 66% on BecomingLit. A static 3DGS reconstruction probe further confirms that our generations are the most multi-view consistent. Built on a video prior, VDAs produce effects beyond the reach of explicit renderers, including physically plausible hair dynamics, shadows, reflections, and refractions.
Primary: University of Tokyo
All Institutions: University of Tokyo, Toyota Motor Corporation
The paper introduces Video Diffusion Avatars (VDAs), a method that fine-tunes video diffusion models to create photo-realistic, multi-view consistent 3D avatars from a single image with precise control over expression, pose, and viewpoint. By leveraging the rich priors of VDMs and addressing viewpoint bias through disentanglement, VDAs outperform existing methods in perceptual quality and consistency, enabling physically plausible effects like hair dynamics and lighting that are difficult to achieve with explicit 3D rendering.
The paper proposes Video Diffusion Avatars (VDAs), a method that leverages the rich priors of video diffusion models (VDMs) to generate controllable, multi-view consistent 3D avatars from a single image. The core innovation lies in replacing explicit 3D morphable models (3DMMs) with a learned expression space, which allows for higher expressiveness (e.g., tongue and mouth interior) but introduces viewpoint bias. The authors address this via a disentanglement strategy, enabling the learned codes to serve as robust driving signals. Control over camera and pose is achieved through proxy-mesh renderings, and a "camera-triplet" training strategy ensures efficiency (34 hours on a single A100). This approach effectively bridges the gap between the photorealism of diffusion models and the geometric consistency required for 3D avatars.
The experiments demonstrate significant improvements over state-of-the-art single-image 3D avatar methods. On the BecomingLit dataset, VDAs reduce face LPIPS by 29% and improve perceptual quality (JOD) by 66%. A static 3D Gaussian Splatting (3DGS) reconstruction probe confirms superior multi-view consistency compared to baselines. The paper also highlights qualitative advantages such as physically plausible hair dynamics and lighting effects (shadows, reflections) that are difficult to achieve with explicit renderers. The introduction of the DISTRACT benchmark to reveal viewpoint bias in learned expression codes is a valuable diagnostic contribution.
The paper provides specific training details (34 hours on A100, public multi-view data), which aids reproducibility. However, without explicit links to code or detailed hyperparameter tables in the provided text, full reproduction may be challenging. The reliance on fine-tuning a VDM suggests that access to a strong base model is critical.
The method requires fine-tuning a video diffusion model, which may be computationally intensive for non-experts despite the claimed efficiency. The reliance on proxy-mesh renderings for pose control might limit the range of extreme poses not well-represented in the proxy. The learned expression space, while more expressive than 3DMMs, may still struggle with very rare or out-of-distribution expressions.
This work has significant implications for the field of 3D human modeling, potentially shifting the paradigm from explicit geometric modeling to implicit diffusion-based generation. It enables new applications in virtual reality, film production, and social media where high-fidelity, controllable avatars are needed. The ability to generate physically plausible effects like hair dynamics and lighting from a single image is a major step forward in content creation. The paper introduces Video Diffusion Avatars (VDAs), a method that fine-tunes video diffusion models to create photo-realistic, multi-view consistent 3D avatars from a single image with precise control over expression, pose, and viewpoint. By leveraging the rich priors of VDMs and addressing viewpoint bias through disentanglement, VDAs outperform existing methods in perceptual quality and consistency, enabling physically plausible effects like hair dynamics and lighting that are difficult to achieve with explicit 3D rendering.
Preference optimization can improve every ranking metric while building a hidden, shared representation shift that the relative loss barely sees, one that can eventually destroy generation. We trace this failure to the common-mode offset, a shared directional shift in the model's hidden states that appears across token positions, prompts, and both chosen and rejected responses. Because it moves chosen and rejected representations together, the relative ranking loss barely reacts to its accumulation. As this offset grows, it projects onto the output layer to form a vocabulary well, elevating a single token's probability until it dominates decoding and causes generation collapse. The isolated offset alone reconstructs the vocabulary well, injecting the learned offset into an untreated base model recreates the collapse where norm-matched random directions do not, and masking this common component from the loss prevents the well in our flagship setting. Under the per-token ranking loss we study, the same collapse appears in 14 of 17 base-model lineages and under full-parameter fine-tuning. Moreover, the offset is not merely an artifact of collapse, since it is already present in textbook DPO and SimPO runs that form no well. Textbook DPO keeps it weakly aligned with any single vocabulary direction even with three times our base budget. Trained from base models for that budget, however, full-parameter SimPO drives it into collapse on both Llama-2 and Llama-3.1, and the same kind of offset grows, without a well, from a public SFT checkpoint under the official SimPO recipe. Across 17 UltraFeedback runs, the better a run ranks held-out responses, the less often two public reward models prefer its samples to the base model's. These findings expose a blind spot in standard ranking metrics and introduce a representation-level signal that can warn of generation collapse before it occurs.
Primary: The University of Tokyo
All Institutions: The University of Tokyo, TIER IV, Inc.
The paper identifies a hidden "common-mode offset" in preference optimization that causes generation collapse despite improved ranking metrics, providing a causal explanation and effective mitigation strategy. This is a significant contribution to the understanding of LLM alignment failures, offering both a diagnostic signal and a practical solution that is likely to influence future alignment research and practice.
The paper introduces a novel diagnostic framework for preference optimization failures, identifying a "common-mode offset" in the final hidden states as the root cause of vocabulary collapse. The methodology is rigorous, combining theoretical analysis of loss invariances with extensive empirical validation. The authors define the offset as the mean displacement of hidden states between trained and base models, demonstrating that this shared vector moves chosen and rejected representations in tandem, thereby evading the relative ranking loss. They propose a "dose" metric to quantify the risk of collapse and introduce a mitigation strategy involving centering the rejected-side loss and applying norm/loudness penalties to the offset. The causal interventions (injecting the offset into base models, masking it from the loss) are well-designed and effectively isolate the mechanism.
The experimental scope is impressive, covering 17 base-model lineages (0.5B to 14B parameters) across five tokenizer families, using both LoRA and full-parameter fine-tuning. The use of two distinct datasets (driving preferences and UltraFeedback) and multiple objectives (DPO, SimPO, Plackett-Luce ranking) strengthens the generalizability of the findings. The discovery that ranking metrics improve while generation collapses (anticorrelation with external reward models) is a significant empirical contribution. The cross-family validation of the mitigation strategy up to three times the base budget is convincing.
The paper provides detailed descriptions of training configurations, measurement protocols, and penalty terms. The authors state that code and data will be released upon publication. The specific hyperparameters (learning rates, batch sizes, LoRA ranks) are clearly defined. The measurement of "dose" and "loudness" is well-defined, allowing for replication. The reliance on specific public checkpoints (e.g., Llama-3-8B SFT from SimPO authors) aids reproducibility.
The mitigation strategy is calibrated on specific datasets and may require tuning for other domains. The analysis is primarily focused on the final hidden state, and the interaction with earlier layers is not deeply explored. The "diffuse degradation" observed in SimPO is noted but not fully resolved by the proposed offset control, indicating a second failure mode that remains unaddressed. The study is limited to English-centric models and datasets.
This work has high impact for the LLM alignment community by exposing a critical blind spot in standard evaluation metrics. It provides a practical diagnostic tool (monitoring the common-mode offset) that can be integrated into training pipelines to prevent generation collapse. The findings challenge the assumption that improving ranking metrics guarantees healthy generation, prompting a re-evaluation of alignment objectives. The proposed penalties offer a simple, effective remedy for a widespread issue. The paper identifies a hidden "common-mode offset" in preference optimization that causes generation collapse despite improved ranking metrics, providing a causal explanation and effective mitigation strategy. This is a significant contribution to the understanding of LLM alignment failures, offering both a diagnostic signal and a practical solution that is likely to influence future alignment research and practice.
Autonomous agents can increasingly propose ideas, modify code, and run experiments, creating an opportunity to accelerate scientific discovery using teams of collaborative AI researchers. To escape local minima in the discovery, we propose to encourage orthogonal exploration by constructing the team using agents with varying expertise. However, realizing this design requires more than diverse backgrounds. When experiments are costly, a team can test only a limited number of ideas, thereby creating the need to balance exploration and exploitation when selecting a subset of researchers to propose and execute new experiments. Furthermore, each result should shape what the team tries next. We therefore ask how AI researchers with different scientific backgrounds can build on one another's findings and sustain progress under a limited experimental budget. To address these challenges, we build a team of persistent AI researchers, each grounded in a different discipline, who keep private research histories, share experimental evidence, and receive research opportunities through adaptive allocation. We demonstrate the effectiveness of this design by deploying the team to search for linear attention architectures, where it outperforms GDN2 at 1.3B scale on language modeling, common-sense reasoning, and synthetic and real-world retrieval tasks. In controlled comparisons, removing any single organizational component from our design yields markedly weaker discoveries: without disciplinary backgrounds, the search plateaus early; without private histories, researchers make less progress despite identical shared evidence; and uniform selection falls short of adaptive allocation within the same budget. These results suggest that sustained autonomous discovery depends not only on stronger agents, but also on who explores, what they remember, and how research effort is allocated.
Primary: University of California, Santa Barbara (UCSB)
All Institutions: University of California, Santa Barbara (UCSB)
The paper introduces a multi-agent team structure with diverse disciplinary backgrounds, private memories, and adaptive budget allocation to autonomously discover linear attention architectures. By demonstrating that this organizational design outperforms ablated versions and achieves competitive results against state-of-the-art linear attention models at 1.3B scale, it provides a rigorous framework for collaborative AI research that balances exploration and exploitation, offering a scalable path toward autonomous scientific discovery in machine learning architecture design.
The paper proposes a multi-agent framework for autonomous architecture search, specifically targeting linear attention mechanisms. The core methodological contribution is the "organizational" structure of the agents: (1) Diverse disciplinary grounding to encourage orthogonal exploration, (2) Private research histories to maintain individual context, and (3) Adaptive allocation of experimental budget based on past performance. This moves beyond simple prompt engineering or single-agent loops by introducing a team dynamic that balances exploration and exploitation. The approach is conceptually sound and addresses a known limitation in LLM-based research agents (local minima due to homogeneous reasoning patterns).
The experiments are rigorous in scope, scaling from 150M to 1.3B parameters. The claim that the discovered architecture outperforms GDN2 (Gated DeltaNet) at 1.3B scale on language modeling, reasoning, and retrieval is significant if true. The ablation studies are critical to the paper's validity: removing disciplinary backgrounds, private histories, or adaptive allocation all lead to worse performance. This isolates the contribution of the team structure from the underlying LLM capabilities. However, the baseline comparison is limited to GDN2; comparisons with other recent linear attention variants (e.g., Mamba, RWKV, or standard Transformers at the same FLOP/param budget) would strengthen the claim.
The paper includes a detailed reproducibility statement, promising appendices with model configurations, hyperparameters, and agent trajectories. The use of FLOP-matched budget allocation is a strong signal of methodological rigor. However, without access to the actual code and agent prompts, full verification is difficult. The reliance on "persistent" agents implies a specific state-management infrastructure that must be clearly documented.
The primary limitation is the high computational cost of the search process, which may limit adoption by smaller labs. Additionally, the generalizability of the "disciplinary background" approach to other domains (e.g., computer vision or reinforcement learning) is not tested. The paper focuses heavily on linear attention, and it is unclear if the team structure yields similar gains in other architectural spaces. There is also a risk of "reward hacking" where agents optimize for the specific evaluation metrics rather than true generalization.
This work contributes to the emerging field of AI-driven scientific discovery. If the organizational principles (diversity, memory, adaptive allocation) are validated, they could be applied to other automated research tasks, such as hyperparameter tuning, dataset curation, or even hypothesis generation in biology. The potential to accelerate the discovery of efficient attention mechanisms is high, given the current industry focus on reducing inference costs. The paper introduces a multi-agent team structure with diverse disciplinary backgrounds, private memories, and adaptive budget allocation to autonomously discover linear attention architectures. By demonstrating that this organizational design outperforms ablated versions and achieves competitive results against state-of-the-art linear attention models at 1.3B scale, it provides a rigorous framework for collaborative AI research that balances exploration and exploitation, offering a scalable path toward autonomous scientific discovery in machine learning architecture design.
Large Language Model (LLM) agents can handle increasingly complex tasks, navigating long trajectories of observations and actions. As these trajectories grow, key-value (KV) cache I/O and storage become the dominant cost. Trajectories can also exceed the LLM's maximum context length, requiring control over active cache growth. Prior work uses context engineering to reduce cache size by trimming or summarizing history. However, modifying the trajectory prefix invalidates subsequent cached states and triggers costly re-prefilling of retained context. Task accuracy can also suffer when too little context is retained to limit re-prefill costs. We propose cache engineering as a new scheme. It allows agents to directly control their active KV cache without modifying the trajectory or re-prefilling retained context. Specifically, agents can proactively use drop to exclude selected cache from future attention and repos to adjust retained keys' position encodings. We formalize these cache operations and identify history dependence as a serving challenge: identical token sequences can yield different cache states under different cache operation sequences, making prefix matching insufficient for correct cache reuse. Thus, we co-design the AKV serving engine. It provides a stateless interface compatible with existing serving APIs, while internally managing stateful cache reuse, eviction, and restoration. Experiments show that a tool-response rolling-drop cache policy improves average benchmark scores by 1.7 points and reduces average cost by 27.7%, compared with conventional threshold-triggered summarization. With AKV, it achieves up to 1.2x the output throughput and a 1.5x speedup in P95 request latency. Recursive self-improvement of cache policies further reduces cost by 7.1% over expert design. Our code is at https://github.com/thu-nics/AKV.
Primary: Tsinghua University
All Institutions: Tsinghua University
The paper proposes cache engineering as a new scheme for LLM agents, enabling direct control over KV cache without trajectory modification, and introduces the AKV serving engine to handle stateful cache reuse. This represents a significant technical contribution to LLM inference optimization, offering substantial improvements in cost and latency for agent-based applications.
The paper introduces "Cache Engineering," a paradigm shift from context engineering (trimming/summarizing) to direct manipulation of the Key-Value (KV) cache. The core method involves two operations: `drop` (excluding specific cache entries from future attention) and `repos` (adjusting position encodings of retained keys). The authors formalize the challenge of "history dependence," where identical token sequences can result in different cache states depending on prior operations, breaking standard prefix-caching assumptions. To address this, they propose the AKV serving engine, which maintains a stateless external interface but manages stateful cache reuse, eviction, and restoration internally. The methodology is sound and addresses a critical bottleneck in long-horizon agent tasks.
Experiments demonstrate that a tool-response rolling-drop policy improves average benchmark scores by 1.7 points while reducing costs by 27.7% compared to summarization. The AKV engine achieves 1.2x output throughput and 1.5x P95 latency speedup. The inclusion of recursive self-improvement for cache policies, yielding an additional 7.1% cost reduction, is a strong empirical result. The evaluation covers both accuracy and efficiency metrics, providing a comprehensive view of the trade-offs.
The code is available at the provided GitHub repository. The paper provides sufficient detail on the cache operations and serving engine design to allow for reproduction. The use of standard benchmarks and clear metric definitions enhances reproducibility.
The approach relies on specific modifications to the attention mechanism and position encodings, which may not be directly compatible with all LLM architectures or inference frameworks without significant adaptation. The "history dependence" issue, while solved by AKV, adds complexity to the serving infrastructure. The paper focuses on text-based agents; applicability to multimodal agents with different cache structures is not explored.
This work has significant implications for the deployment of LLM agents in production environments where cost and latency are critical. By decoupling cache management from context modification, it enables more flexible and efficient agent designs. The AKV engine could become a standard component in LLM serving stacks, particularly for applications involving long trajectories. The paper proposes cache engineering as a new scheme for LLM agents, enabling direct control over KV cache without trajectory modification, and introduces the AKV serving engine to handle stateful cache reuse. This represents a significant technical contribution to LLM inference optimization, offering substantial improvements in cost and latency for agent-based applications.
Long-context language models (LCLMs) power agents that act over long trajectories, from improving a codebase to running research experiments to forecasters that condition on lengthy histories. These applications draw on three distinct long-context capabilities: recall, where the answer is a fact present in the context; state tracking, where the outcome can be computed by performing a series of transformations to a specified state; and pattern learning, where the goal is to make a prediction by sampling from a distribution inferred from many observations. Existing benchmarks primarily focus on the first capability. In this work, we introduce LEAPS (Long-context Evaluation of Adaptive Pattern learning and State tracking), a collection of 12 novel datasets, spanning agentic workflows, behavioral modeling, and forecasting of market and world events. By construction, LEAPS is resistant to data contamination, extends to millions of input tokens, and uses reliable metrics. We evaluate 30 API and open models from 8K to 512K input tokens. LEAPS distinguishes models that perform near perfectly on recall tasks, finding that such models' state tracking and pattern learning abilities can vary significantly and degrade much more dramatically with increasing input lengths--even for pattern learning, where more observations should make learning a theoretically easier task. Moreover, thinking models often underperform their base counterparts, exhausting their generation budget or oversimplifying the task. We hope LEAPS can guide the development of the next generation of LCLMs.
Primary: Jane Street
All Institutions: Jane Street, Princeton University
The paper introduces LEAPS, a comprehensive benchmark for evaluating long-context language models on recall, state tracking, and pattern learning. It provides rigorous experimental evidence that current models struggle with state tracking and pattern learning at scale, offering critical insights for the development of next-generation LCLMs.
The paper introduces LEAPS, a benchmark suite comprising 12 datasets designed to evaluate three specific long-context capabilities: recall, state tracking, and pattern learning. The methodology is rigorous, emphasizing resistance to data contamination through procedural generation of synthetic data and the use of reliable metrics. The distinction between these three capabilities is a strong conceptual contribution, as existing benchmarks often conflate them or focus heavily on recall. The approach to extending evaluations to millions of tokens is methodologically sound, though the reliance on API models for the upper end of the scale introduces potential variability in evaluation conditions.
The experimental scope is extensive, evaluating 30 models (both API and open-source) across input lengths from 8K to 512K tokens. The results provide valuable insights, particularly the finding that models excelling in recall do not necessarily perform well in state tracking or pattern learning. The observation that "thinking" models often underperform their base counterparts due to budget exhaustion or oversimplification is a significant and counter-intuitive empirical finding. The degradation of performance with increasing context length, even in tasks where more data should theoretically help (pattern learning), is a critical result for the field.
The authors provide a strong reproducibility statement, releasing code for data generation and evaluation, as well as the original evaluation data in Parquet format (with exceptions for licensing). The use of synthetic data generation scripts ensures that the benchmark can be regenerated, further mitigating contamination risks. The exclusion of social media data due to licensing is a minor limitation but does not significantly impact the core reproducibility of the main findings.
The primary limitation is the reliance on API models for the longest context lengths, which may not be fully reproducible by other researchers due to cost and access constraints. Additionally, the benchmark focuses on specific types of long-context tasks (agentic, behavioral, forecasting), which may not cover all potential applications of LCLMs. The performance of open models at the 512K token limit may be limited by hardware constraints, potentially skewing the comparison if not all models can be tested at the maximum length.
This benchmark has high potential impact as it addresses a critical gap in the evaluation of long-context models. By distinguishing between recall, state tracking, and pattern learning, it provides a more nuanced view of model capabilities than existing benchmarks. The findings regarding the limitations of current models, particularly the degradation with length and the underperformance of thinking models, will likely guide future research directions in model architecture and training. The benchmark's resistance to contamination makes it a robust tool for ongoing evaluation as models evolve. The paper introduces LEAPS, a comprehensive benchmark for evaluating long-context language models on recall, state tracking, and pattern learning. It provides rigorous experimental evidence that current models struggle with state tracking and pattern learning at scale, offering critical insights for the development of next-generation LCLMs.
Human video scales easily and does not need robot hardware, but existing methods require depth sensing, overlay the robot onto 2D video, or generate data only for the demonstrated objects. We present Real2Sem2Real (R2Sem2R), a real-to-semantic-to-real pipeline that turns one monocular RGB video of a human demonstration into thousands of robot demonstrations across the semantic object instances of a category. R2Sem2R tracks the SE(3) pose of each object with DINOv3-distilled Gaussian splats, outperforming existing methods and FoundationPose with estimated metric depth. R2Sem2R then aligns digital cousins, each generated from one or a few photographs, to the demonstrated objects with a semantic alignment. This alignment matches whole objects and the functional regions of target objects, such as a coffee machine's tray. R2Sem2R maps the human hand to a feasible robot grasp that extends to the cousins. Robot demonstrations are generated either in IsaacSim, for photorealistic rendering, or MuJoCo, for contact physics, and R2Sem2R can generate over 2,000 episodes in about two hours. On two real robots and four tasks, we train a vision-language-action policy on data from one human video per task, and policies trained on our data match or exceed policies trained on teleoperation across diverse object instances. The project page is available at https://real2sem2real.github.io/.
Primary: Stanford University
All Institutions: Stanford University, ETH Zurich
The paper presents a novel and effective pipeline for generating semantically augmented robot demonstrations from a single human RGB video. By combining robust RGB-only object tracking with semantic alignment of digital cousins, it enables the transfer of functional knowledge across object instances, resulting in policies that match or exceed those trained on teleoperation with significantly less human effort.
The paper proposes a coherent "Real-to-Semantic-to-Real" pipeline that addresses a critical bottleneck in robot learning: the cost of collecting diverse, object-specific demonstrations. The core technical contribution is a robust RGB-only 4D object tracker using DINOv3-distilled Gaussian splats, which outperforms depth-dependent baselines like FoundationPose. The novelty lies in the "semantic alignment" of digital cousins (generated from few photos) to the demonstrated objects, allowing the transfer of functional affordances (e.g., where to place a mug on a specific machine) rather than just geometric similarity. The integration of hand-to-gripper retargeting and parallel simulation in IsaacSim/MuJoCo is well-engineered, though the reliance on VLMs for keyframe selection and region segmentation introduces potential fragility.
The experimental setup is rigorous, involving two real robots (xArm7, Panda) and four distinct tasks that test different aspects of the pipeline (placement, pouring, hanging, reorientation). The comparison against teleoperation is the gold standard, and the results showing R2Sem2R matching or exceeding teleoperation with significantly less human effort are compelling. The ablation study isolating the effect of semantic augmentation (R2Sem2R vs. R2R2R+) is particularly valuable, clearly demonstrating the added value of the cousin alignment. The tracking accuracy benchmarks on YCBInEOAT provide strong evidence for the superiority of the proposed tracker.
The paper provides detailed descriptions of the pipeline components, including loss functions, initialization strategies, and simulation settings. The use of standard tools (IsaacSim, MuJoCo, DINOv3) aids reproducibility. However, the specific implementation details of the "semantic ICP" and the VLM prompts used for region selection are not fully detailed in the main text, which could hinder exact replication. The code availability is implied by the project page, but the paper itself is a preprint/submission, so full code release status is pending.
The method is limited to rigid objects and requires a metric measurement of the demonstrated objects, which is a non-trivial assumption. The IsaacSim kinematic replay lacks contact dynamics, limiting applicability to contact-rich tasks, though MuJoCo is used as an alternative. The pipeline is computationally intensive (2.5 hours for 2000 episodes) and requires careful setup of the scene and camera calibration. The reliance on monocular depth estimation for some components (though mitigated by the splat tracker) remains a potential source of error.
This work has significant potential to democratize robot learning by reducing the barrier to entry for data collection. By enabling the generation of thousands of diverse demonstrations from a single human video, it could accelerate the development of general-purpose robot policies. The approach is applicable to a wide range of manipulation tasks and could be extended to other embodiments and object categories. The paper presents a novel and effective pipeline for generating semantically augmented robot demonstrations from a single human RGB video. By combining robust RGB-only object tracking with semantic alignment of digital cousins, it enables the transfer of functional knowledge across object instances, resulting in policies that match or exceed those trained on teleoperation with significantly less human effort.
Despite recent advances in policy learning, long-horizon manipulation remains difficult because learned policies must avoid compounding errors while preserving future feasibility. While search-based planning can explicitly reason over such future consequences, it quickly becomes expensive in high-dimensional continuous action spaces. This raises a central question: how can long-horizon reasoning from search be transferred into policies in continuous action spaces? To this end, we propose DISCO: Distilled Tree Search over Continuous Actions. DISCO uses progressive-widening MCTS in physics simulation to generate long-horizon plans. During search, a learned policy proposes continuous actions, while a learned value function scores actions by their expected progress toward task completion. Using GPU-parallelized simulation, DISCO periodically distills MCTS-generated plans into the policy-value model, improving the action proposals and value estimates used in subsequent searches. Finally, the learned policy-value model can be directly deployed without the need for simulator-based search at test time. Across simulated and real-world rearrangement tasks, DISCO outperforms behavior cloning, reinforcement learning, and search-based baselines.
Primary: Carnegie Mellon University
All Institutions: Carnegie Mellon University, Stanford University
The paper introduces DISCO, a method that distills long-horizon reasoning from GPU-parallelized MCTS into a deployable policy-value model for continuous action spaces. It represents a strong contribution to robot manipulation by effectively bridging the gap between sample-efficient planning and scalable policy learning, offering a viable solution to the compounding error problem in long-horizon tasks.
The paper proposes DISCO, a framework that bridges the gap between sample-efficient search-based planning and scalable policy learning in continuous action spaces. The core innovation is the use of progressive-widening Monte Carlo Tree Search (MCTS) within a physics simulator, where a learned policy proposes actions and a learned value function scores them. Crucially, the system employs a distillation loop: MCTS-generated plans are periodically distilled back into the policy-value model. This iterative refinement allows the policy to internalize the long-horizon reasoning capabilities of the search process. The method relies on GPU-parallelized simulation to make the search tractable, which is a significant engineering contribution. The final deployed model is a standard policy-value network that operates without the simulator, offering a clear advantage in real-world deployment latency and robustness compared to online search methods.
The evaluation covers both simulated and real-world multi-object rearrangement tasks. The paper claims to outperform behavior cloning (BC), reinforcement learning (RL), and search-based baselines. The inclusion of real-world experiments is a strong positive, as many papers in this domain remain purely simulated. The comparison against search-based baselines is particularly relevant, as it isolates the benefit of the distillation process over raw search performance. However, the specific metrics for "long-horizon" success rates and the degree of error compounding reduction would need to be scrutinized in the full results section to confirm the magnitude of improvement.
The method relies on specific hyperparameters for the MCTS (progressive widening factors, exploration constants) and the distillation frequency. The use of GPU-parallelized simulation suggests a dependency on specific hardware or simulation frameworks (likely Isaac Gym or similar), which may limit immediate reproducibility for labs without such infrastructure. The project page is provided, which likely contains code, aiding reproducibility.
The primary limitation is the dependency on a high-fidelity physics simulator for the training phase. If the simulator-to-real-world gap is significant, the distilled policy may suffer from performance degradation, although the real-world results suggest this is mitigated. Additionally, the computational cost of GPU-parallelized MCTS during training can be substantial, potentially limiting the scalability to extremely high-dimensional action spaces or very long horizons beyond what was tested.
This work has significant implications for the field of robot learning, particularly for tasks requiring precise, long-horizon manipulation. By demonstrating that search-based reasoning can be effectively distilled into a reactive policy, it offers a pathway to combining the robustness of planning with the speed of learned policies. This could influence future approaches to hierarchical reinforcement learning and model-based control. The paper introduces DISCO, a method that distills long-horizon reasoning from GPU-parallelized MCTS into a deployable policy-value model for continuous action spaces. It represents a strong contribution to robot manipulation by effectively bridging the gap between sample-efficient planning and scalable policy learning, offering a viable solution to the compounding error problem in long-horizon tasks.
Humanoid robots operating in human-centric environments must coordinate the entire body to traverse tight, cluttered spaces while maintaining balance. Existing methods often struggle to jointly consider dynamic stability and collision avoidance from high-dimensional perceptual inputs. Moreover, for humanoids, safe behavior is inherently multimodal: the same obstacle may admit multiple valid traversals, which is difficult to represent with a single deterministic policy. To address these challenges, we propose SWAP (Steerable Whole-body Avoidance Policy), a hierarchical framework that separates collision-aware decision making from stabilization. A high-level, perception-conditioned diffusion policy generates multimodal, collision-aware motion proposals, while a low-level whole-body RL controller tracks these proposals and maintains dynamic balance. To efficiently train the high-level policy, we obtain scalable supervision via hindsight labeling: we roll out the low-level policy under randomized commands, place obstacles only in the space that each executed motion did not sweep, so the motion is collision-free by construction, and re-render the perception the robot would have seen. At deployment, Monte Carlo sampling with simple cost signals steers the diffusion model toward safe, task-aligned behaviors without heavy reward tuning. Experiments in simulation and on a real Unitree G1 humanoid show reliable avoidance with static and dynamic obstacles. Test-time steering further enables customization, improving targeted rewards (e.g., speed, orientation, or posture) by 34% on average while maintaining a high collision avoidance rate. Together, these results support a practical solution to humanoid whole-body collision avoidance with test-time steerability. Project website: https://iscoyizj.github.io/SWAP-steerable-humanoid-oa-website/
Primary: Carnegie Mellon University
All Institutions: Carnegie Mellon University, Amazon FAR
The paper presents a novel hierarchical framework for humanoid whole-body collision avoidance that effectively combines diffusion-based multimodal planning with RL-based stabilization, demonstrating strong real-world performance and test-time steerability. The use of hindsight labeling for efficient training and the separation of collision-aware decision-making from dynamic balance represent significant methodological contributions that address key challenges in safe humanoid navigation.
The paper proposes SWAP, a hierarchical framework for humanoid whole-body collision avoidance that decouples high-level decision-making from low-level stabilization. The high-level component utilizes a perception-conditioned diffusion policy to generate multimodal motion proposals, addressing the inherent multimodality of safe navigation (e.g., going left vs. right around an obstacle). A key methodological contribution is the use of hindsight labeling for training the high-level policy: by rolling out a low-level RL controller under randomized commands and placing obstacles only in the space *not* swept by the robot, the authors generate collision-free supervision signals without requiring a complex reward function or simulator-in-the-loop collision detection during training. This is a clever and efficient training strategy. At deployment, Monte Carlo sampling with simple cost signals steers the diffusion model, allowing for test-time customization of behavior (e.g., prioritizing speed or posture) without retraining. The separation of concerns is well-motivated, and the use of diffusion models for multimodal action generation in robotics is a timely and relevant architectural choice.
The experiments are conducted in simulation and on a real Unitree G1 humanoid robot, which is a strong and practical choice for demonstrating real-world applicability. The evaluation includes both static and dynamic obstacles, covering a range of challenging scenarios. The paper reports a 34% average improvement in targeted rewards (speed, orientation, posture) via test-time steering while maintaining a high collision avoidance rate. While the specific metrics for "collision avoidance rate" and the baseline comparisons are not detailed in the provided text, the combination of simulation and real-robot validation, along with the quantitative improvement in steerability, suggests a rigorous experimental setup. The ability to demonstrate results on a physical humanoid is a significant strength, as many robotics papers remain purely in simulation.
The paper provides a project website, which likely contains code, videos, and additional details. The use of a standard platform (Unitree G1) and common simulation tools (implied) enhances reproducibility. The hindsight labeling technique is clearly described, allowing other researchers to implement similar training strategies. However, without access to the full code and hyperparameters, full reproducibility is difficult to assess. The reliance on a specific robot model may limit immediate reproducibility for labs without access to Unitree hardware, but the simulation components should be replicable.
The primary limitation is the reliance on a specific humanoid robot (Unitree G1) for real-world validation, which may not generalize directly to other platforms without significant adaptation. The method involves a two-stage process (high-level diffusion policy + low-level RL controller), which increases computational complexity and implementation difficulty compared to end-to-end approaches. The test-time steering via Monte Carlo sampling may introduce latency, which could be a concern in high-speed dynamic environments. Additionally, the paper does not extensively discuss failure modes or the limits of the approach in extremely cluttered or unpredictable environments.
This work has significant potential impact on the field of humanoid robotics and safe robot navigation. The hierarchical approach combining diffusion policies for multimodal planning with RL for stabilization is a promising paradigm that could be adapted to other robotic systems. The hindsight labeling technique offers a scalable way to generate training data for collision avoidance, which is a common challenge in robotics. The test-time steerability feature is particularly valuable for human-centric environments where user preferences or task-specific requirements may vary. This work could influence future research on safe robot operation, multimodal policy learning, and the integration of generative models in robotics. The paper presents a novel hierarchical framework for humanoid whole-body collision avoidance that effectively combines diffusion-based multimodal planning with RL-based stabilization, demonstrating strong real-world performance and test-time steerability. The use of hindsight labeling for efficient training and the separation of collision-aware decision-making from dynamic balance represent significant methodological contributions that address key challenges in safe humanoid navigation.
Video models provide rich priors for generalizable robotic manipulation, but plausible visual futures are not always executable. We introduce DreamCan, a hierarchical policy that grounds video models through sparse visual planning: a high-level planner adapted from a pre-trained video model and a low-level visuomotor policy. Rather than predicting dense trajectories, the high-level planner generates sparse visual subgoals that balance reachability and task utility. From offline, action-free robot videos, we estimate these quantities and use dynamic programming to select subgoal sequences without manual annotations. The resulting subgoals are more reliably attained than uniformly spaced alternatives, and the full system outperforms language, dense-video, and uniform-subgoal planners, reaching high planning success rates on Language-Table tasks and 91.2% success on LIBERO-Long with a 0.1B-parameter low-level policy. Additional suboptimal videos and test-time compute further improve success. The system also performs challenging real-world bimanual tasks, including box folding and packing with unseen object combinations. More results are available at https://dreamcan.github.io/.
Primary: Stanford University
All Institutions: Stanford University, Harvard University, National University of Singapore, Northwestern University
DreamCan introduces a hierarchical policy that uses sparse visual subgoals from video models to enable long-horizon robot manipulation. The paper presents a rigorous evaluation showing that this approach outperforms dense and uniform planning baselines, achieving high success rates in simulation and real-world bimanual tasks, thereby demonstrating the viability of video-based sparse planning for complex robotic manipulation.
The paper proposes "DreamCan," a hierarchical policy that bridges the gap between generative video models and executable robot actions. The core innovation lies in the high-level planner, which is adapted from a pre-trained video model to generate *sparse* visual subgoals rather than dense trajectories. This is a significant architectural choice, as dense video predictions are often physically infeasible for current robot hardware. The method estimates subgoal reachability and task utility from offline, action-free videos, utilizing dynamic programming to select optimal subgoal sequences without manual annotation. This unsupervised discovery of subgoals is a strong methodological contribution, reducing the reliance on expensive human demonstrations for long-horizon tasks. The low-level policy is a goal-conditioned visuomotor policy that executes the actions to reach these subgoals.
The experimental evaluation is robust, covering both simulation (Language-Table, LIBERO-Long) and real-world bimanual tasks. Achieving 91.2% success on LIBERO-Long with a small 0.1B-parameter low-level policy is a notable result, suggesting that the sparse planning approach effectively offloads complexity from the low-level controller. The comparison against language-based, dense-video, and uniform-subgoal planners provides a comprehensive baseline analysis. The inclusion of real-world bimanual tasks (box folding, packing) with unseen object combinations demonstrates generalization capabilities beyond simulation. The observation that additional suboptimal videos and test-time compute improve success rates is a valuable empirical finding for practitioners.
The paper provides a project URL (dreamcan.github.io) which likely contains code and additional results. The methodology relies on pre-trained video models and standard dynamic programming, which are well-understood components. However, the specific adaptation of the video model for sparse subgoal generation and the estimation of reachability/utility from action-free videos may require careful implementation details that are not fully visible in the abstract. The use of offline data is a practical advantage for reproducibility in real-world settings where collecting new action data is costly.
The reliance on pre-trained video models means the system's performance is bounded by the quality and coverage of the video priors. If the video dataset does not contain examples of the specific task or object configurations, the planner may fail. The method is designed for manipulation tasks; its applicability to locomotion or other robotic domains is not discussed. The computational cost of the high-level planner (video model inference) and the dynamic programming step could be a bottleneck for real-time operation, although test-time compute scaling is mentioned as a positive factor.
This work has significant implications for the field of robot learning, particularly for long-horizon manipulation tasks. By leveraging video models as priors for planning, it offers a path towards more generalizable and data-efficient robot policies. The sparse subgoal approach is a practical solution to the "plausibility vs. executability" problem in video-based planning. This could influence future research in hierarchical reinforcement learning and the integration of generative models into robotic control loops. DreamCan introduces a hierarchical policy that uses sparse visual subgoals from video models to enable long-horizon robot manipulation. The paper presents a rigorous evaluation showing that this approach outperforms dense and uniform planning baselines, achieving high success rates in simulation and real-world bimanual tasks, thereby demonstrating the viability of video-based sparse planning for complex robotic manipulation.
Differences in hand geometry, joint parameterization, and kinematic topology pose challenges for cross-embodiment dexterous grasp generation. However, spatial relations among links alone are insufficient to fully capture the hand's kinematic topology. We propose DexGraspFlow, a conditional link-level $\mathrm{SE}(3)$ flow-matching framework integrating geometric relations and kinematic topology. The framework encodes local object geometry, link geometry, and current poses. A dual-stream DiT models link--link coordination and link--object interaction, while attention biases derived from kinematic-tree distances provide explicit structural information for modeling evolving geometric interactions. The model jointly learns translational and rotational velocity fields, integrates them to generate link poses, and recovers joint configurations through inverse kinematics. Additionally, inference-time palm-orientation guidance enables controllable grasp generation without retraining. Across three robotic hands and 30 object--hand pairs, DexGraspFlow achieves a 95.33\% average simulation success rate, with an average generation latency, including inverse kinematics, approximately 49\% lower than that of a state-of-the-art baseline. Further experiments demonstrate zero-shot adaptability to seven finger-removal variants and palm-orientation control. A model trained on demonstrations for the deployed hand achieves a 91\% success rate over 100 real-robot grasping trials, supporting the framework's applicability to physical grasp execution.
Primary: Peking University
All Institutions: Peking University, University of Science and Technology of China
DexGraspFlow introduces a topology-aware SE(3) flow matching framework that significantly advances cross-embodiment dexterous grasp generation by explicitly encoding kinematic structure, achieving state-of-the-art success rates and efficiency with strong real-world validation.
The paper proposes DexGraspFlow, a framework for cross-embodiment dexterous grasp generation using SE(3) flow matching. The core methodological contribution is the integration of kinematic topology (via URDF-derived shortest-path distances) into the attention mechanism of a dual-stream DiT architecture. This addresses a specific gap in prior work (like TRO-Grasp) where spatial proximity was used as a proxy for kinematic relevance, which fails across different hand morphologies. The use of flow matching over diffusion models is a timely choice, offering faster inference (10 steps vs. 50+ for diffusion) and stable training. The formulation of separate translational (Euclidean) and rotational (SO(3) geodesic) velocity fields is mathematically sound and preserves the manifold structure of rotations. The inference-time palm-orientation guidance via tangent-space velocity corrections is a clever, training-free extension that enhances controllability.
The experimental evaluation is rigorous and comprehensive. The authors benchmark against strong, recent baselines (TRO-Grasp, DRO-Grasp, GenDexGrasp) across three distinct robotic hands (Barrett, Allegro, ShadowHand) and 30 object-hand pairs. The success rate of 95.33% is state-of-the-art, outperforming the previous best (TRO-Grasp) by a significant margin. The latency improvement (~49% lower) is substantial and practically important. The zero-shot generalization to finger-removal variants is a strong test of the topology-aware representation, showing the model understands the structural changes. The real-robot experiments (91% success rate) provide crucial validation beyond simulation, although the sample size (100 trials) is modest. The ablation studies clearly demonstrate the necessity of both geometric and topology biases.
The paper provides detailed implementation specifics, including dataset splits, hyperparameters (learning rate, batch size, network dimensions), and evaluation protocols (Isaac Gym settings, force magnitudes). The use of standard datasets (CMapDataset, ContactDB, YCB) and open-source tools (Pyroki, Isaac Gym) enhances reproducibility. However, the code and model weights are not explicitly linked in the provided text (though typically available for such papers), and the specific "filtered CMapDataset split" might require clarification if not publicly released.
The primary limitation is the reliance on a fixed set of training hands; generalization to completely novel hand morphologies (e.g., a 3-fingered hand vs. 5-fingered) is not tested, only finger removals. The method does not explicitly enforce joint limits or collision avoidance during the flow generation, relying on post-hoc IK which can fail or produce unnatural poses if the generated link poses are infeasible. The real-robot evaluation is limited to a single hand type (Linker Hand) and a small number of objects.
This work has significant implications for the field of dexterous manipulation. By providing a fast, accurate, and topology-aware method for cross-embodiment grasp generation, it facilitates the transfer of manipulation skills across different robotic platforms. The flow matching approach is likely to become a standard for generative robotics tasks due to its efficiency. The training-free guidance mechanism offers a practical way to add controllability to generative models without retraining, a valuable technique for other robotic applications. DexGraspFlow introduces a topology-aware SE(3) flow matching framework that significantly advances cross-embodiment dexterous grasp generation by explicitly encoding kinematic structure, achieving state-of-the-art success rates and efficiency with strong real-world validation.
Humanoid loco-manipulation is particularly challenging because it requires coordinating locomotion, balance, and object interaction, making it difficult to learn from limited demonstrations or from scratch through Reinforcement Learning (RL). We introduce DiPLO (self-distilled Diffusion Planning for Loco-manipulation), a framework that combines a self-distilled diffusion planner with a low-level whole-body tracking policy. Our planner is pretrained via supervised learning on motions derived from human-to-humanoid retargeting, while the tracker is trained via RL to execute them. DiPLO centers around two complementary post-training stages. First, we introduce Self-Distillation Diffusion Policy Optimisation (SD:DPO) to remove the planner's dependence on strong task conditioning, which we identify as a crucial ingredient for effective pretraining, as downstream fine-tuning does not recover comparable performance when it is omitted.. We then fine-tune this distilled model with Diffusion-Augmented Proximal Policy Optimization (DA:PPO), a maximum-entropy RL method for diffusion models, to improve task success and align generated plans with the frozen tracker's execution capabilities. Together, these stages enable motion generation through state-history and goal conditioning alone, enabling 99% task success in pick-carry-place tasks, compared to just 45% for the pretrained model. The resulting planner also generates motions for unseen object goals, with further evaluations on kicking and pushing/pulling tasks. Finally, we deploy our pipeline on the hardware, using our DA:PPO planner-tracker on a Unitree G1, achieving repeated successful executions without any real-world fine-tuning.
Primary: Technical University of Munich (TUM)
All Institutions: Technical University of Munich (TUM), McGill University, INRIA, Institute for Advanced Study (IAS)
DiPLO introduces a self-distilled diffusion planning framework that enables humanoid robots to perform complex loco-manipulation tasks from limited demonstrations by decoupling motion generation from execution and using a two-stage post-training process to remove dependence on kinematic references. The method demonstrates strong generalization to unseen goals and successful real-world deployment, highlighting the potential of hierarchical learning approaches for embodied AI.
The paper proposes DiPLO, a hierarchical framework for humanoid loco-manipulation that decouples high-level motion planning (diffusion model) from low-level execution (RL tracking policy). The core methodological contribution is a two-stage post-training process: (1) Self-Distillation Diffusion Policy Optimization (SD:DPO), which distills a teacher model conditioned on kinematic references into a student model conditioned only on state history and goals, effectively removing the need for explicit task priors at deployment; and (2) Diffusion-Augmented Proximal Policy Optimization (DA:PPO), which fine-tunes the distilled planner using maximum-entropy RL to align generated plans with the frozen tracker's capabilities. The use of SO(3) diffusion for rotational features is a technically sound choice that respects the manifold geometry of rotations. The separation of planning and tracking allows the planner to learn diverse motion structures from limited data while the tracker learns robust feedback control via RL, addressing the mode-covering vs. mode-seeking trade-off inherent in generative models vs. RL.
The experimental evaluation is rigorous and comprehensive. The authors provide extensive ablations on pretraining data augmentation (SBTO vs. symmetry), the importance of task conditioning during pretraining, and the specific benefits of SO(3) diffusion. They compare motion-level planning against action-level diffusion baselines (PDP, BeyondMimic), demonstrating the superiority of the hierarchical approach for survival and task success. The generalization evaluation, where the robot achieves unseen goals up to 5 meters away, is a strong indicator of the method's robustness. The real-world deployment on a Unitree G1 humanoid, achieving 100% success in pick-carry-place tasks without real-world fine-tuning, provides strong evidence of practical viability.
The paper provides detailed descriptions of the diffusion processes (VP diffusion for Euclidean features, IG_SO(3) for rotations), the DA-MDP formulation, and the training objectives. However, specific hyperparameters, network architectures (DiT details), and code availability are not explicitly provided in the text (no GitHub link found in the provided text). The reliance on specific datasets (OmniRetarget, OMOMO) and simulation environments (mjlab) may limit immediate reproducibility for those without access to these resources.
The method relies on external motion capture for object pose estimation in the real-world setup, which is a significant practical limitation for general deployment. The planning frequency is limited to ~12.5 Hz due to diffusion sampling, which may be too slow for highly dynamic tasks requiring faster replanning. The evaluation is primarily focused on pick-carry-place and similar tasks; performance on more complex, multi-object, or highly dynamic interactions is not fully explored. The dependence on a frozen tracker means that if the tracker fails, the planner cannot adapt its strategy in real-time to compensate for execution errors beyond the replanning horizon.
This work contributes to the growing field of humanoid robotics by providing a scalable method for learning complex loco-manipulation skills from limited demonstrations. The insight that strong task conditioning is useful for pretraining but can be distilled away for deployment is a valuable principle for robot learning. The combination of diffusion models for planning and RL for tracking offers a robust paradigm for handling the non-stationarity and balance constraints inherent in humanoid control. DiPLO introduces a self-distilled diffusion planning framework that enables humanoid robots to perform complex loco-manipulation tasks from limited demonstrations by decoupling motion generation from execution and using a two-stage post-training process to remove dependence on kinematic references. The method demonstrates strong generalization to unseen goals and successful real-world deployment, highlighting the potential of hierarchical learning approaches for embodied AI.
Every robot already carries high-resolution contact sensors: its own joints. Legged robots and humanoids exploit them to walk, run, and track whole-body motions without vision, but there the contacts are large and forceful. Is joint sensing also enough for precise, contact-rich assembly, where the robot must find a millimeter-scale opening, mesh gear teeth, or catch a thread through light contact? We study this question by training blind assembly policies that observe only a short history of joint positions, velocities, and their own actions, without visual, force, torque, or tactile sensing. The poses of both the fixture and the part are unknown to the policy, and the grasp is randomized substantially. The policy must therefore assemble the parts through contact, sensed exclusively through its own proprioception. Trained with reinforcement learning entirely in simulation over randomized parts, grasps, and contacts, such policies learn peg insertion, gear meshing, and nut threading. The peg insertion policy transfers zero-shot to the real world and inserts pegs of different shapes and sizes in 73% of trials. Controlled comparisons with matched visual policies in simulation show that blind policies generalize far better to unseen part shapes and sizes, though they can be slower. Training such policies needs thousands of different assemblies, each simulated with its clearance intact, so we also release an open-source generator that builds them procedurally from exact convex pieces for MuJoCo; it is more accurate and faster than convex decomposition or MuJoCo's signed-distance collision. Videos and interactive simulations are at https://younghyopark.me/blind-proprioceptive-assembly/.
Primary: Massachusetts Institute of Technology
All Institutions: Massachusetts Institute of Technology
The paper demonstrates that proprioceptive sensing alone is sufficient for precise, contact-rich robotic assembly tasks like peg insertion, gear meshing, and nut threading, achieving zero-shot sim-to-real transfer and superior generalization to unseen parts compared to visual policies. By introducing a rigorous simulation framework with exact convex collision geometry and a novel contact-supervised RL training method, the authors provide a robust, sensor-light alternative to vision-based manipulation, significantly advancing the state of the art in blind dexterity and sim-to-real transfer for precision tasks.
The paper proposes a novel approach to precise robotic assembly using only proprioceptive sensing (joint positions, velocities, and actions), explicitly excluding vision, force, torque, or tactile sensors. The core method involves training Reinforcement Learning (PPO) policies in simulation with extensive domain randomization of part geometry, grasp, and contact properties. A key technical contribution is the "contact supervision" auxiliary loss, which trains the policy's latent representation to predict contact states (free, touching, aligned) without providing privileged information to the actor, thereby accelerating learning from sparse rewards. Additionally, the authors introduce "AssemblyGen," an open-source procedural generator for MuJoCo that creates exact convex collision geometries for tight-clearance assemblies (pegs, gears, nuts), addressing a significant simulation bottleneck where standard convex decomposition or SDF methods fail to preserve sub-millimeter clearances accurately.
The experimental evaluation is rigorous and comprehensive. The authors demonstrate that blind policies achieve near-perfect success rates on in-distribution tasks and, crucially, generalize significantly better to unseen part shapes and sizes compared to matched visual policies. The zero-shot sim-to-real transfer is a strong result, with the peg insertion policy achieving 73% success on real hardware without fine-tuning, including on out-of-distribution pegs. The ablation studies effectively isolate the contribution of the contact supervision loss and the collision geometry generator. The comparison against scripted spiral search baselines highlights the efficiency of the learned reactive search strategy.
The paper provides high reproducibility standards. It releases an open-source generator for the simulation assets, which is a major barrier in this field. Detailed hyperparameters, architecture descriptions, and training ranges are provided in the appendix. The use of standard simulators (MuJoCo) and common algorithms (PPO) further aids reproducibility. The specific hardware setup (Flexiv Rizon 4S, Robotiq 2F-85) is clearly defined.
The primary limitation is the assumption that the robot starts in contact with the fixture; the policy does not handle the initial approach phase from a distance. The behavior is local and cannot recover from large deviations (e.g., if the part slips out of the grasp). The real-world evaluation is limited to peg insertion; gear meshing and nut threading were only validated in simulation. The reliance on precise simulation of friction and contact dynamics, despite domain randomization, remains a potential source of sim-to-real gap for other tasks.
This work challenges the prevailing assumption that precise manipulation requires vision or force sensing. By demonstrating that proprioception alone is sufficient for millimeter-scale assembly, it opens new avenues for robust, low-cost robotic systems that do not require expensive sensors or complex calibration. The released simulation tools will likely benefit the broader robotics community working on contact-rich tasks. The finding that blind policies generalize better to unseen geometries than visual policies is a significant insight for the design of robust manipulation policies. The paper demonstrates that proprioceptive sensing alone is sufficient for precise, contact-rich robotic assembly tasks like peg insertion, gear meshing, and nut threading, achieving zero-shot sim-to-real transfer and superior generalization to unseen parts compared to visual policies. By introducing a rigorous simulation framework with exact convex collision geometry and a novel contact-supervised RL training method, the authors provide a robust, sensor-light alternative to vision-based manipulation, significantly advancing the state of the art in blind dexterity and sim-to-real transfer for precision tasks.
Consensus is a fundamental problem in distributed computing and at the core of state machine replication (SMR). Despite decades of active human research, discovering practical low-latency solutions to the problem is still open, as it requires combining protocol knowledge, correctness reasoning, implementation, and experimental evaluation. We report on how AI helped us discover A-greement, a new consensus protocol for geo-replicated state machines that significantly outperforms state-of-the-art solutions. Averaged across six deployments with 5 or 9 replicas, A-greement reduces the mean latency of a geo-distributed key-value store over the fastest baseline by 12.2% and 33.3% on write-heavy and read-heavy workloads, respectively, and matches the fastest baseline on read-only workloads. A-greement combines lease-authorized local reads, direct client learning for leader-ordered writes, and a fast fixed-quorum per-key write path, selected according to network topology and workload. At the heart of our discovery lies AI-greement, an agent harness for the "autonomous" discovery, implementation, experimentation, and verification of candidate protocols. AI-greement integrates (1) a concept-oriented protocol knowledge base, (2) a modular implementation skeleton, and (3) a Lean-based verification framework with reusable safety definitions, reference specifications, and proofs. AI-greement is of independent interest: its Lean formalization uncovers several previously unreported vulnerabilities in existing consensus protocols, and it provides guidelines on how domain-specific harnesses can turn general-purpose coding agents into effective tools for (distributed) computing research. Yet we argue that human consensus experts remain essential before, during, and after the search: agents reshape the researcher's role rather than end consensus research.
Primary: École Polytechnique Fédérale de Lausanne (EPFL)
All Institutions: École Polytechnique Fédérale de Lausanne (EPFL)
The paper presents AI-greement, a specialized agent harness that leverages formal verification and domain knowledge to autonomously discover and verify a new consensus protocol, A-greement, which outperforms state-of-the-art solutions in geo-distributed settings. This work is significant as it demonstrates a novel application of AI agents in the design of complex distributed systems, moving beyond code generation to algorithmic discovery and formal verification, thereby offering a new paradigm for tackling hard problems in distributed computing.
The paper proposes "AI-greement," a specialized agent harness designed to autonomously discover, implement, experiment with, and verify distributed consensus protocols. The methodology integrates three key components: a concept-oriented protocol knowledge base, a modular implementation skeleton, and a Lean-based formal verification framework. This approach moves beyond simple code generation by embedding domain-specific constraints and formal safety definitions directly into the agent's workflow. The resulting protocol, A-greement, is a hybrid design combining lease-authorized local reads, direct client learning for leader-ordered writes, and fast fixed-quorum per-key write paths. The use of Lean for verification is a significant methodological strength, as it allows the agent to reason about correctness rather than just performance, leading to the discovery of previously unreported vulnerabilities in existing protocols.
The evaluation is rigorous, comparing A-greement against state-of-the-art baselines across six different geo-distributed deployments with 5 and 9 replicas. The results show a 12.2% reduction in mean latency for write-heavy workloads and a 33.3% reduction for read-heavy workloads compared to the fastest baseline. The paper also validates the safety of the discovered protocol using the Lean framework. The inclusion of real-world network topologies and varied workload characteristics strengthens the empirical claims.
The paper provides details on the agent harness architecture and the Lean verification framework, which are critical for reproducibility. However, as the full text provided is a skeleton with section headers, specific implementation details, prompt engineering strategies, and code repositories are not visible in the excerpt. The mention of a "modular implementation skeleton" suggests that code artifacts are likely available or planned, but without the actual code or detailed prompt logs, full reproducibility of the *discovery process* is difficult to assess from the abstract and headers alone.
The primary limitation is the reliance on a highly specialized, domain-specific harness (AI-greement) which may not generalize easily to other areas of distributed computing or software engineering without significant adaptation. Additionally, the "autonomous" nature of the discovery is tempered by the authors' admission that human experts remain essential, suggesting the current level of autonomy is limited to search space reduction and verification assistance rather than fully independent invention. The performance gains, while significant, are specific to geo-replicated state machines and may not apply to other consensus contexts.
This paper has high potential impact on the field of AI for Systems. It demonstrates that LLMs can be effectively used not just for coding, but for the *design* of complex distributed algorithms when coupled with formal verification tools. The discovery of vulnerabilities in existing protocols via the Lean framework highlights the utility of AI-assisted formal methods. This work could inspire similar "agent harnesses" for other hard problems in computer science where correctness and performance are tightly coupled. The paper presents AI-greement, a specialized agent harness that leverages formal verification and domain knowledge to autonomously discover and verify a new consensus protocol, A-greement, which outperforms state-of-the-art solutions in geo-distributed settings. This work is significant as it demonstrates a novel application of AI agents in the design of complex distributed systems, moving beyond code generation to algorithmic discovery and formal verification, thereby offering a new paradigm for tackling hard problems in distributed computing.