论文检索

Optimization · Everything Else

Kaheon Kim, Bohan Zhou, Changbo Zhu, Xiaohui Chen

This paper introduces a new constraint-free concave dual formulation for the Wasserstein barycenter. Tailoring the vanilla dual gradient ascent algorithm to the Sobolev geometry, we derive a scalable Sobolev gradient ascent (SGA) algorithm to compute the barycenter for input distributions supported on a regular grid. Despite the algorithmic simplicity, we provide a global convergence analysis that achieves the same rate as the classical subgradient descent methods for minimizing nonsmooth convex functions in the Euclidean space. A central feature of our SGA algorithm is that the computationally expensive $c$-concavity projection operator enforced on the Kantorovich dual potentials is unnecessary to guarantee convergence, leading to significant algorithmic and theoretical simplifications over all existing primal and dual methods for computing the exact barycenter. Our numerical experiments demonstrate the superior empirical performance of SGA over the existing optimal transport barycenter solvers.

Reinforcement Learning · Deep RL

Enrico Marchesini, Eva Boguslawski, Alessandro Leite, Christopher Amato, Matthieu DUSSARTRE, Marc Schoenauer, Benjamin Donnot, Priya Donti

Improving power grid operations is essential for enhancing flexibility and accelerating grid decarbonization. Reinforcement learning (RL) has shown promise in this domain, most notably through the Learning to Run a Power Network (L2RPN) competition series, but prior work has primarily focused on single-agent settings, neglecting the often decentralized, multi-agent nature of grid control. We fill this gap with MARL2Grid-TR, the first multi-agent RL (MARL) benchmark for grid topology and redispatching, developed in collaboration with transmission system operators. Built on RTE France’s high-fidelity simulation platform, our benchmark supports decentralized control across substations and generators, with configurable agent scopes, observability settings, expert-informed heuristics, and safety-critical constraints. The benchmark includes a suite of realistic scenarios that expose key challenges, such as coordination under partial information, long-horizon objectives, and adherence to hard physical constraints. Empirical results show that current MARL methods struggle under these real-world conditions. By providing a standardized, extensible platform, we aim to advance the development of scalable, cooperative, and safe learning algorithms for power grids.

Reinforcement Learning · Deep RL

Shenao Zhang, Donghan Yu, Yihao Feng, Bowen Jin, Zhaoran Wang, John Peebles, Zirui Wang

Large language models excel with reinforcement learning (RL), but fully unlocking this potential requires a mid-training stage. Intuitively, an effective mid-training stage should both learn a strong policy prior and enable fast learning through online interactions. We formalize this intuition by presenting the first theoretical result on how mid-training shapes post-training: it acquires strong policy priors by efficiently pruning the action space and accelerates RL convergence by shortening the effective planning horizon. Moreover, we prove that temporal abstractions simultaneously compress the size of the action set and reduce the decision horizon, thereby improving regret minimization after training. Building on these insights, we introduce Reasoning as Action Abstractions (RA3), a scalable mid-training algorithm. Specifically, we derive a temporal variational bound and optimize it by iteratively discovering temporally-consistent latent structures via RL, then fine-tuning on the bootstrapped data. Experiments on code generation tasks demonstrate the effectiveness of our approach. Across multiple base models, RA3 improves the average performance on HumanEval and MBPP by 8 and 4 points over the base model and the next-token prediction baseline. Furthermore, RA3 achieves faster convergence and higher asymptotic performance in RLVR on HumanEval+, MBPP+, LiveCodeBench, and Codeforces.

Deep Learning · Everything Else

Md Zesun Ahmed Mia, Malyaban Bal, Abhronil Sengupta

The quadratic complexity of self-attention mechanism presents a significant impediment to applying Transformer models to long sequences. This work explores computational principles derived from astrocytes—glial cells critical for biological memory and synaptic modulation—as a complementary approach to conventional architectural modifications for efficient self-attention. We introduce the Recurrent Memory Augmented Astromorphic Transformer (RMAAT), an architecture integrating abstracted astrocyte functionalities. RMAAT employs a recurrent, segment-based processing strategy where persistent memory tokens propagate contextual information. An adaptive compression mechanism, governed by a novel retention factor derived from simulated astrocyte long-term plasticity (LTP), modulates these tokens. Attention within segments utilizes an efficient, linear-complexity mechanism inspired by astrocyte short-term plasticity (STP). Training is performed using Astrocytic Memory Replay Backpropagation (AMRB), a novel algorithm designed for memory efficiency in recurrent networks. Evaluations on the Long Range Arena (LRA) benchmark demonstrate RMAAT's competitive accuracy and substantial improvements in computational and memory efficiency, indicating the potential of incorporating astrocyte-inspired dynamics into scalable sequence models.

Reinforcement Learning · Online

Klemens Iten, Lenart Treven, Bhavya, Florian Dorfler, Andreas Krause

Reinforcement learning algorithms are typically designed for discrete-time dynamics, even though the underlying real-world control systems are often continuous in time. In this paper, we study the problem of continuous-time reinforcement learning, where the unknown system dynamics are represented using nonlinear ordinary differential equations (ODEs). We leverage probabilistic models, such as Gaussian processes and Bayesian neural networks, to learn an uncertainty-aware model of the underlying ODE. Our algorithm, COMBRL, greedily maximizes a weighted sum of the extrinsic reward and model epistemic uncertainty. This yields a scalable and sample-efficient approach to continuous-time model-based RL. We show that COMBRL achieves sublinear regret in the reward-driven setting, and in the unsupervised RL setting (i.e., without extrinsic rewards), we provide a sample complexity bound. In our experiments, we evaluate COMBRL in both standard and unsupervised RL settings and demonstrate that it scales better, is more sample-efficient than prior methods, and outperforms baselines across several deep RL tasks.

Deep Learning · Everything Else

Haojie Duanmu, Size Zheng, Ningxin Zheng, Jianqiao Lu, Xuegui Zheng, Xingcheng Zhang, Li-Wen Chang, Xin Liu, Dahua Lin

The quadratic complexity of softmax attention poses a major bottleneck for long-context modeling, motivating a surge of linear attention variants with linear complexity. Unlike softmax attention, which benefits from optimized kernels, linear attention lacks general-purpose, hardware-efficient support and scalable distributed implementations. We introduce **Flex**ible **L**inear **A**ttention (FlexLA), a domain-specific compiler that automates the generation of high-performance, scalable kernels for a wide range of linear attention models directly from high-level PyTorch code. At its core, FlexLA employs an intuitive programming abstraction that decomposes any linear attention algorithm into three canonical phases: intra-chunk computation, inter-chunk state propagation, and output merging. This unified abstraction enables FlexLA to perform domain-specific optimizations, automatically generating kernels that fuse computation and communication at a fine-grained tile level and eliminating host synchronization. Our evaluation demonstrates that FlexLA combines programmability with performance: a wide range of linear attention variants can be implemented in just a few dozen lines of code, while the generated kernels deliver 1.01x-4.9x the performance of sate-of-the-art expert-optimized library and scale with near-linear efficiency on scalar gated linear attention to 16 million tokens on 128 GPUs, surpassing the state-of-the-art distributed baseline by up to 7.2x.

Applications · Robotics

Muleilan Pei, Shaoshuai Shi, Shaojie Shen

Scalable and realistic simulation of multi-agent traffic behavior is critical for advancing autonomous driving technologies. Although existing data-driven simulators have made significant strides in this domain, they predominantly rely on supervised learning to align simulated distributions with real-world driving scenarios. A persistent challenge, however, lies in the distributional shift that arises between training and testing, which often undermines model generalization in unseen environments. To address this limitation, we propose SMART-R1, a novel R1-style reinforcement fine-tuning paradigm tailored for next-token prediction models to better align agent behavior with human preferences and evaluation metrics. Our approach introduces a metric-oriented policy optimization algorithm to improve distribution alignment and an iterative "SFT-RFT-SFT" training strategy that alternates between Supervised Fine-Tuning (SFT) and Reinforcement Fine-Tuning (RFT) to maximize performance gains. Extensive experiments on the large-scale Waymo Open Motion Dataset (WOMD) validate the effectiveness of this simple yet powerful R1-style training framework in enhancing foundation models. The results on the Waymo Open Sim Agents Challenge (WOSAC) showcase that SMART-R1 achieves state-of-the-art performance with an overall realism meta score of 0.7858, ranking first on the leaderboard at the time of submission.

Reinforcement Learning · Batch/Offline

Natinael Neggatu, Jeremie Houssineau, Giovanni Montana

Most existing offline RL methods presume the availability of action labels within the dataset, but in many practical scenarios, actions may be missing due to privacy, storage, or sensor limitations. We formalise the setting of action-free offline-to-online RL, where agents must learn from datasets consisting solely of $(s,r,s')$ tuples and later leverage this knowledge during online interaction. To address this challenge, we propose learning state policies that recommend desirable next-state transitions rather than actions. Our contributions are twofold. First, we introduce a simple yet novel state discretisation transformation and propose Offline State-Only DecQN (OSO-DecQN), a value-based algorithm designed to pre-train state policies from action-free data. OSO-DecQN integrates the transformation to scale efficiently to high-dimensional problems while avoiding instability and overfitting associated with continuous state prediction. Second, we propose a novel mechanism for guided online learning that leverages these pre-trained state policies to accelerate the learning of online agents. Together, these components establish a scalable and practical framework for leveraging action-free datasets to accelerate online RL. Empirical results across diverse benchmarks demonstrate that our approach improves convergence speed and asymptotic performance, while analyses reveal that discretisation and regularisation are critical to its effectiveness.

Sadia Asif, Mohammad Mohammadi Amiri

Large language models deployed in sensitive applications increasingly require the ability to unlearn specific knowledge, such as user requests, copyrighted materi- als, or outdated information, without retraining from scratch to ensure regulatory compliance, user privacy, and safety. This task, known as machine unlearning, aims to remove the influence of targeted data (forgetting) while maintaining per- formance on the remaining data (retention). A common approach is to formu- late this as a multi-objective problem and reduce it to a single-objective prob- lem via scalarization, where forgetting and retention losses are combined using a weighted sum. However, this often results in unstable training dynamics and degraded model utility due to conflicting gradient directions. To address these challenges, we propose OFMU, a penalty-based bi-level optimization framework that explicitly prioritizes forgetting while preserving retention through a hierar- chical structure. Our method enforces forgetting via an inner maximization step that incorporates a similarity-aware penalty to decorrelate the gradients of the for- get and retention objectives, and restores utility through an outer minimization step. To ensure scalability, we develop a two-loop algorithm with provable conver- gence guarantees under both convex and non-convex regimes. We further provide a rigorous theoretical analysis of convergence rates and show that our approach achieves better trade-offs between forgetting efficacy and model utility compared to prior methods. Extensive experiments across vision and language benchmarks demonstrate that OFMU consistently outperforms existing unlearning methods in both forgetting efficacy and retained utility.

Deep Learning · Algorithms

Donghyun Lee, Yuhang Li, Ruokai Yin, Shiting Xiao, Priyadarshini Panda

State Space Models (SSMs) have emerged as powerful alternatives to attention-based Transformers, with Mamba demonstrating impressive efficiency and scalability. As these models grow increasingly larger, the need for Parameter-Efficient Fine-Tuning (PEFT) methods becomes critical to adapt pre-trained Mamba to downstream tasks without prohibitive computational costs. However, previous approaches simply apply traditional Transformer-tailored PEFT methods without addressing the unique temporal processing dynamics of SSMs. To address this limitation, we propose ***Memba***, a membrane-driven PEFT approach specifically designed for Mamba. ***Memba*** introduces Leaky Integrate Membrane (LIM) neurons as bio-inspired gating mechanisms that naturally accumulate membrane potentials over time, enhancing selective information retention. By strategically combining LIM neurons with Low-Rank Adaptations (LoRA) and cross-layer membrane transfer, our approach significantly improves Mamba's temporal modeling capabilities. Extensive experiments across language and vision tasks demonstrate that ***Memba*** achieves substantial improvements over existing PEFT methods.

Computer Vision · Image and Video Generation

Anirud Aggarwal, Abhinav Shrivastava, Matthew Gwilliam

Diffusion-based image generation models excel at producing high-quality synthetic content, but suffer from slow and computationally expensive inference. Prior work has attempted to mitigate this by caching and reusing features within diffusion transformers across inference steps. These methods, however, often rely on rigid heuristics that result in limited acceleration or poor generalization across architectures. We propose **E**volutionary **C**aching to **A**ccelerate **D**iffusion models (ECAD), a genetic algorithm that learns efficient, per-model, caching schedules forming a Pareto frontier, using only a small set of calibration prompts. ECAD requires no modifications to network parameters or reference images. It offers significant inference speedups, enables fine-grained control over the quality-latency trade-off, and adapts seamlessly to different diffusion models. Notably, ECAD's learned schedules can generalize effectively to resolutions and model variants not seen during calibration. We evaluate ECAD on PixArt-alpha, PixArt-Sigma, and FLUX.1-dev using multiple metrics (FID, CLIP, Image Reward) across diverse benchmarks (COCO, MJHQ-30k, PartiPrompts), demonstrating consistent improvements over previous approaches. On PixArt-alpha, ECAD identifies a schedule that outperforms the previous state-of-the-art method by 4.47 COCO FID while increasing inference speedup from 2.35x to 2.58x. Our results establish ECAD as a scalable and generalizable approach for accelerating diffusion inference. Our project page and code are available here: https://research.aniaggarwal.com/ecad

Deep Learning · Algorithms

Yifan Zhu, 慧强 荣, Haoran Luo

Large Language Models (LLMs) often hallucinate, generating content inconsistent with the input. Retrieval-Augmented Generation (RAG) and Reinforcement Learning with Human Feedback (RLHF) can mitigate hallucinations but require resource-intensive retrieval or large-scale fine-tuning. Decoding-based methods are lighter yet lack explicit hallucination control. To address this, we present \textbf{Token-Guard}, a token-level hallucination control method based on self-checking decoding. Token-Guard performs internal verification at each reasoning step to detect hallucinated tokens before they propagate. Candidate fragments are further evaluated in a latent space with explicit hallucination risk scoring, while iterative pruning and regeneration dynamically correct detected errors. Experiments on HALU datasets show Token-Guard substantially reduces hallucinations and improves generation accuracy, offering a scalable, lightweight solution for reliable LLM outputs. Our code is publicly available\footnote{\url{https://github.com/rhq945/Token-Guard}}.

Reinforcement Learning · Deep RL

Ying-Tu Chen, Wei Hung, Bing-Shu Wu, Zhang-Wei Hong, Ping-Chun Hsieh

Many sequential decision-making tasks involve optimizing multiple conflicting objectives, requiring policies that adapt to different user preferences. In multi-objective reinforcement learning (MORL), one widely studied approach addresses this by training a single policy network conditioned on preference-weighted rewards. In this paper, we explore a novel algorithmic perspective: leveraging reward-free reinforcement learning (RFRL) for MORL. While RFRL has historically been studied independently of MORL, it learns optimal policies for any possible reward function, making it a natural fit for MORL's challenge of handling unknown user preferences. We propose using the RFRL's training objective as an auxiliary task to enhance MORL, enabling more effective knowledge sharing beyond the multi-objective reward function given at training time. To this end, we adapt a state-of-the-art RFRL algorithm to the MORL setting and introduce a preference-guided exploration strategy that focuses learning on relevant parts of the environment. Through extensive experiments and ablation studies, we demonstrate that our approach significantly outperforms the state-of-the-art MORL methods across diverse MO-Gymnasium tasks, achieving superior performance and data efficiency. This work provides the first systematic adaptation of RFRL to MORL, demonstrating its potential as a scalable and empirically effective solution to multi-objective policy learning.

Applications · Robotics

Patrick Yin, Tyler Westenbroek, Zhengyu Zhang, Ignacio Dagnino, Eeshani Shilamkar, Numfor Mbiziwo-Tiapo, Simran Bagaria, Xinlei Liu, Galen Mullins, Andrey Kolobov 等

Reinforcement learning in massively parallel physics simulations has driven major progress in sim-to-real robot learning. However, current approaches remain brittle and task-specific, relying on extensive per-task engineering to design rewards, curricula, and demonstrations. Even with this engineering, they often fail on long-horizon, contact-rich manipulation tasks and do not meaningfully scale with compute, as performance quickly saturates when training revisits the same narrow regions of state space. We introduce \Method, a simple and scalable framework that enables on-policy reinforcement learning to robustly solve a broad class of dexterous manipulation tasks using a single reward function, fixed algorithm hyperparameters, no curricula, and no human demonstrations. Our key insight is that long-horizon exploration can be dramatically simplified by using simulator resets to systematically expose the RL algorithm to the diverse set of robot-object interactions which underlie dexterous manipulation. \Method\ programmatically generates such resets with minimal human input, converting additional compute directly into broader behavioral coverage and continued performance gains. We show that \Method\ gracefully scales to long-horizon dexterous manipulation tasks beyond the capabilities of existing approaches and is able to learn robust policies over significantly wider ranges of initial conditions than baselines. Finally, we distill \Method \ into visuomotor policies which display robust retrying behavior and substantially higher success rates than baselines when transferred to the real world zero-shot. Project webpage: https://omnireset.github.io

Optimization · Large Scale, Parallel and Distributed

Rohit Jena, Vedant Zope, Pratik A Chaudhari, James Gee

In this work, we propose FFDP, a set of IO-aware non-GEMM fused kernels supplemented with a distributed framework for image registration at unprecedented scales. Image registration is an inverse problem fundamental to biomedical and life sciences, but algorithms have not scaled in tandem with image acquisition capabilities. Our framework complements existing model parallelism techniques proposed for large-scale transformer training by optimizing non-GEMM bottlenecks and enabling convolution-aware tensor sharding. We demonstrate unprecedented capabilities by performing multimodal registration of a 100μm ex-vivo human brain MRI volume at native resolution – an inverse problem more than 570× larger than a standard clinical datum in about a minute using only 8 A6000 GPUs. FFDP accelerates existing state-of-the-art optimization and deep learning registration pipelines by upto 6 − 7× while reducing peak memory consumption by 20 − 59%. Comparative analysis on a 250μm dataset shows that FFDP can fit upto 64× larger problems than existing SOTA on a single GPU, and highlights both the performance and efficiency gains of FFDP compared to SOTA image registration methods.

Deep Learning · Generative Models and Autoencoders

Jacob Bamberger, Iolo Jones, Dennis Duncan, Michael Bronstein, Pierre Vandergheynst, Adam Gosztolai

Deep generative models often face a fundamental tradeoff: high sample quality can come at the cost of memorisation, where the model reproduces training data rather than generalising across the underlying data geometry. We introduce Carré du champ flow matching (CDC-FM), a generalisation of flow matching (FM), that improves the quality-generalisation tradeoff by regularising the probability path with a geometry-aware noise. Our method replaces the homogeneous, isotropic noise in FM with a spatially varying, anisotropic Gaussian noise whose covariance captures the local geometry of the latent data manifold. We prove that this geometric noise can be optimally estimated from the data and is scalable to large data. Further, we provide an extensive experimental evaluation on diverse datasets (synthetic manifolds, point clouds, single-cell genomics, animal motion capture, and images) as well as various neural network architectures (MLPs, CNNs, and transformers). We demonstrate that CDC-FM consistently offers a better quality-generalisation tradeoff, even when used as a latent space generation model. We observe significant improvements over standard FM in data-scarce regimes and in highly non-uniformly sampled datasets, which are often encountered in AI for science applications. Our work provides a mathematical framework for studying the interplay between data geometry, generalisation and memorisation in generative models, as well as a robust and scalable algorithm that can be readily integrated into existing flow matching pipelines.

Reinforcement Learning · Deep RL

Devvrit Khatri, Lovish Madaan, Rishabh Tiwari, Rachit Bansal, Venkata Sai Surya Subramanyam Duvvuri, Manzil Zaheer, Inderjit Dhillon, David Brandfonbrener, Rishabh Agarwal

Reinforcement learning (RL) has become central to training large language models (LLMs), yet the field lacks predictive scaling methodologies comparable to those established for pre-training. Despite rapidly rising compute budgets, there is no principled understanding of how to evaluate algorithmic improvements for scaling RL compute. We present the first large-scale systematic study, amounting to more than 400,000 GPU-hours, that defines a principled framework for analyzing and predicting RL scaling in LLMs. We fit sigmoidal compute-performance curves for RL training and ablate a wide range of common design choices to analyze their effects on asymptotic performance and compute efficiency. We observe: (1) Not all recipes yield similar asymptotic performance, Details such as loss aggregation, normalization, curriculum, and off-policy algorithm primarily modulate compute efficiency without materially shifting the asymptote, and (3) Stable, scalable recipes follow predictable scaling trajectories, enabling extrapolation from smaller-scale runs. Combining these insights, we propose a _best-practice_ recipe, ScaleRL, and demonstrate its effectiveness by successfully scaling and predicting validation performance on a single RL run scaled up to 100,000 GPU-hours. Our work provides both a _scientific framework_ for analyzing scaling in RL and a practical recipe that brings RL training closer to the predictability long achieved in pre-training.

Social Aspects · Fairness, Equity, Justice and Safety

Kihyun Kim, Jiawei Zhang, Asuman Ozdaglar, Pablo Parrilo

Conventional preference learning methods often prioritize opinions held more widely when aggregating preferences from multiple evaluators. This may result in policies that are biased in favor of some types of opinions or groups and susceptible to strategic manipulation. To address this issue, we develop a novel preference learning framework capable of aligning aggregate opinions and policies proportionally with the true population distribution of evaluator preferences. Grounded in social choice theory, our approach infers the feasible set of evaluator population distributions directly from pairwise comparison data. Using these estimates, the algorithm constructs a policy that satisfies foundational axioms from social choice theory, namely monotonicity and Pareto efficiency, as well as our newly-introduced axioms of population-proportional alignment and population-bounded manipulability. Moreover, we propose a soft-max relaxation method that smoothly trades off population-proportional alignment with the selection of the Condorcet winner (which beats all other options in pairwise comparisons). Finally, we validate the effectiveness and scalability of our approach through experiments on both tabular recommendation tasks and large language model alignment.

Theory · Optimization

Pankaj Agarwal, Oliver Chubet, Sharath Raghvendra, Keegan Yao

Let $(X,d)$ be a metric space and let $\mu,\nu$ be discrete probability distributions supported on finite point sets $A,B \subseteq X$. For any $p \in [1,\infty]$, the {\it $W_p$-distance} between $\mu$ and $\nu$, $W_p(\mu, \nu)$, is defined as the $p$-th root of the minimum cost of transporting all the probability mass from $\mu$ to $\nu$, where moving a probability mass of $\delta$ from $a \in A$ to $b \in B$ incurs a cost of $\delta d(a,b)^p$. We give a (Las Vegas) randomized algorithm that computes a $(4+\varepsilon)$-approximate $W_p$ optimal-transport (OT) plan in $O(n^2 + (n^{3/2}\varepsilon^{-1}\log n\log\Delta)^{1+o(1)}\log U)$ time with probability at least $1-1/n$, for all $p \in [1,\infty]$, where $\varepsilon > 0$ is an arbitrarily small constant and $\Delta$ is the ratio between the largest and smallest interpoint distances in $A\cup B$. The previous best result achieved an $O(\log n)$-approximation in $O(pn^2)$ time, for constant values of $p$. Our algorithm significantly improves the approximation factor and, importantly, is the first quadratic-time method that extends to the $W_\infty$-distance. In contrast, additive approximation methods such as Sinkhorn are efficient only for constant $p$ and fail to handle $p=\infty$. \changed{Our algorithm also extends to a query model where, for any integer $k > 1$, we give an algorithm that preprocesses $X$ into clusters in $O(n^2+kn^{1+1/k}\log n\log\Delta)$ time, after which a $O(k)$-approximate $W_p$ distance between any two distributions $\mu$ and $\nu$ with $X$ as support can be computed in $(n^{1+1/k}\log n\log\Delta)^{1+o(1)}$ time with probability at most $1-1/n$.} Finally, for $p=\infty$, we show that obtaining a relative approximation factor better than $2$ in $O(n^2)$ time would resolve the long-standing open problem of computing a perfect matching in an arbitrary bipartite graph in quadratic time.

Deep Learning · Algorithms

Moritz Akiya Zanger, Pascal R Van der Vaart, Wendelin Boehmer, Matthijs T. J. Spaan

Uncertainty quantification is a critical aspect of reinforcement learning and deep learning, with numerous applications ranging from efficient exploration and stable offline reinforcement learning to outlier detection in medical diagnostics. The scale of modern neural networks, however, complicates the use of many theoretically well-motivated approaches such as full Bayesian inference. Approximate methods like deep ensembles can provide reliable uncertainty estimates but still remain computationally expensive. In this work, we propose contextual similarity distillation, a novel approach that explicitly estimates the variance of an ensemble of deep neural networks with a single model, without ever learning or evaluating such an ensemble in the first place. Our method builds on the predictable learning dynamics of wide neural networks, governed by the neural tangent kernel, to derive an efficient approximation of the predictive variance of an infinite ensemble. Specifically, we reinterpret the computation of ensemble variance as a supervised regression problem with kernel similarities as regression targets. The resulting model can estimate predictive variance at inference time with a single forward pass, and can make use of unlabeled target-domain data or data augmentations to refine its uncertainty estimates. We empirically validate our method across a variety of out-of-distribution detection benchmarks and sparse-reward reinforcement learning environments. We find that our single-model method performs competitively and sometimes superior to ensemble-based baselines and serves as a reliable signal for efficient exploration. These results, we believe, position contextual similarity distillation as a principled and scalable alternative for uncertainty quantification in reinforcement learning and general deep learning.