Reinforcement Learning

Also known as RL

Follow this topic Ask about this topic

Reinforcement learning studies how an agent can choose actions over time to earn rewards, using experience to improve its decisions. Common approaches include value-based methods such as Q-learning, policy-gradient methods such as PPO, and model-based methods that learn or use a model of the environment; offline and hierarchical methods address learning from fixed data and long tasks. Applications range from Atari and MuJoCo benchmarks to robot control, traffic, games, and language-model agents that use tools or solve multi-step problems. A central challenge is making learning efficient and reliable when rewards are sparse or imperfect, environments change, and success in simulation or benchmarks may not transfer safely to deployment.

Latest papers

1,742 papers237 papers in the last four weeks, up 46% on the four weeks before. 1.4% of all new papers.

Feb 16, 2026

Decoupled Continuous-Time Reinforcement Learning via Hamiltonian Flow

Many real-world control problems, ranging from finance to robotics, evolve in continuous time with non-uniform, event-driven decisions. Standard discrete-time reinforcement learning (RL), based on fixed-step Bellman updates, struggles in this setting: as time gaps shrink, the

Q

-function collapses to the value function

V

, eliminating action ranking. Existing continuous-time methods reintroduce action information via an advantage-rate function

q

. However, they enforce optimality through complicated martingale losses or orthogonality constraints, which are sensitive to the choice of test processes. These approaches entangle

V

and

q

into a large, complex optimization problem that is difficult to train reliably. To address these limitations, we propose a novel decoupled continuous-time actor-critic algorithm with alternating updates:

q

is learned from diffusion generators on

V

, and

V

is updated via a Hamiltonian-based value flow that remains informative under infinitesimal time steps, where standard max/softmax backups fail. Theoretically, we prove rigorous convergence via new probabilistic arguments, sidestepping the challenge that generator-based Hamiltonians lack Bellman-style contraction under the sup-norm. Empirically, our method outperforms prior continuous-time and leading discrete-time baselines across challenging continuous-control benchmarks and a real-world trading task, achieving 21% profit over a single quarter

-

nearly doubling the second-best method.

Minh Nguyen

Google · University of Texas at Austin

Reinforcement Learning Continuous-Time RL RL Control Actor-Critic Methods

Feb 13, 2026

Calculating Mutual Information between a Reward Maximizer and its Environment

An important question in the field of AI is the extent to which successful behaviour requires an internal representation of the world. In this work, we quantify the amount of information an optimal policy provides about the underlying environment. We consider a Controlled Markov Process (CMP) with

n

states and

m

actions, assuming a uniform prior over the space of possible transition dynamics. We prove that observing a deterministic policy that is optimal for any non-constant reward function then conveys exactly

n \log m

bits of information about the environment. Specifically, we show that the mutual information between the environment and the optimal policy is

n \log m

bits. This bound holds across a broad class of objectives, including finite-horizon, infinite-horizon discounted, and time-averaged reward maximization. These findings provide a precise information-theoretic lower bound on the ``implicit world model'' necessary for optimality.

Alfred Harwood, Jose Faustino, Alex Altair

1Dovetail Research · University of Sao Paulo

Markov Decision Processes Reinforcement Learning World Models Information-Theoretic Lower Bounds Mutual Information

Feb 12, 2026

TabSieve: Explicit In-Table Evidence Selection for Tabular Prediction

Tabular prediction can benefit from in-table rows as few-shot evidence, yet existing tabular models typically perform instance-wise inference and LLM-based prompting is often brittle. Models do not consistently leverage relevant rows, and noisy context can degrade performance. To address this challenge, we propose TabSieve, a select-then-predict framework that makes evidence usage explicit and auditable. Given a table and a query row, TabSieve first selects a small set of informative rows as evidence and then predicts the missing target conditioned on the selected evidence. To enable this capability, we construct TabSieve-SFT-40K by synthesizing high-quality reasoning trajectories from 331 real tables using a strong teacher model with strict filtering. Furthermore, we introduce TAB-GRPO, a reinforcement learning recipe that jointly optimizes evidence selection and prediction correctness with separate rewards, and stabilizes mixed regression and classification training via dynamic task-advantage balancing. Experiments on a held-out benchmark of 75 classification and 52 regression tables show that TabSieve consistently improves performance across shot budgets, with average gains of 2.92% on classification and 4.45% on regression over the second-best baseline. Further analysis indicates that TabSieve concentrates more attention on the selected evidence, which improves robustness to noisy context.

Yongyao Wang, Ziqi Miao, Lu Yang +4

Gaoling School of Artificial Intelligence, Renmin University of China · Shanghai AI Laboratory

In-Context Example Selection Reinforcement Learning In-Context Learning Tabular ICL Tabular Prediction

Feb 12, 2026

Partial GFlowNet: Accelerating Convergence in Large State Spaces via Strategic Partitioning

Generative Flow Networks (GFlowNets) have shown promising potential to generate high-scoring candidates with probability proportional to their rewards. As existing GFlowNets freely explore in state space, they encounter significant convergence challenges when scaling to large state spaces. Addressing this issue, this paper proposes to restrict the exploration of actor. A planner is introduced to partition the entire state space into overlapping partial state spaces. Given their limited size, these partial state spaces allow the actor to efficiently identify subregions with higher rewards. A heuristic strategy is introduced to switch partial regions thus preventing the actor from wasting time exploring fully explored or low-reward partial regions. By iteratively exploring these partial state spaces, the actor learns to converge towards the high-reward subregions within the entire state space. Experiments on several widely used datasets demonstrate that \modelname converges faster than existing works on large state spaces. Furthermore, \modelname not only generates candidates with higher rewards but also significantly improves their diversity.

Xuan Yu, Xu Wang, Rui Zhu +2

University of Science and Technology of China (USTC), Hefei, China · Suzhou Institute for Advanced Research, USTC, Suzhou, China · State Key Laboratory of Precision and Intelligent Chemistry, USTC, Hefei, China

Reinforcement Learning Generative Flow Networks RL Exploration RL for Generative Models

Feb 11, 2026

Can We Really Learn One Representation to Optimize All Rewards?

As unsupervised pretraining becomes increasingly ubiquitous in reinforcement learning, a more thorough theoretical understanding of these methods becomes of equal importance to their empirical success. We focus on the setting of unsupervised learning via interaction, where the forward-backward (FB) representation learning serves as a prototypical and popular example. In this paper, we shed light on FB by formally contextualizing the method within a broader class of recent methods that use regression to obtain a low-rank approximation of a successor measure ratio. Our analysis clarifies when FB representations can exist and how the low-rank approximation converges in practice. Building upon the theory, we propose a variant of FB that is both more amenable to theoretical understanding and simpler to optimize in practice. Experiments in didactic settings, as well as in

10

state-based and image-based continuous control domains, demonstrate that our method converges to desired representations with

10^5 \times

smaller errors than FB, achieving

+24\%

improved zero-shot performance on average. We also demonstrate that zero-shot policies inferred by our algorithm provide an efficient initialization if the user prefers further fine-tuning on downstream tasks. Our project website is available at https://chongyi-zheng.github.io/onestep-fb.

Chongyi Zheng, Royina Karegoudra Jayanth, Benjamin Eysenbach

Princeton University

Zero-Shot RL Representation Learning Reinforcement Learning Successor Representations Unsupervised Learning Unsupervised Representation Learning+3

Feb 11, 2026

Patch the Distribution Mismatch: RL Rewriting Agent for Stable Off-Policy SFT

Large language models are commonly adapted to downstream tasks through supervised fine-tuning (SFT), but substantial distribution mismatch between downstream supervision and a model's generation distribution can intensify catastrophic forgetting. Data rewriting offers a data-centric way to narrow this mismatch before SFT. Existing methods, however, typically sample rewrites from a prompt-induced conditional distribution, which need not align with the backbone's natural question-answering generation distribution, and fixed templates can reduce output diversity. We formulate data rewriting as a policy-learning problem and train a lightweight LoRA rewriting policy with reinforcement learning. The policy optimizes question-answering-style distributional alignment and semantic diversity under a hard task-consistency gate, producing verified supervision for downstream SFT. Across three instruction-tuned backbones, the resulting models attain downstream gains broadly comparable to standard SFT while reducing degradation on non-downstream benchmarks in every evaluated setting. Additional experiments on logical reasoning and medical question answering provide preliminary evidence that a rewriting policy can be reused across domains for the same backbone.

Jiacheng Wang, Zhijie Liu, Ping Jian +4

Beijing Institute of Technology, Beijing, China · Tsinghua University, Beijing, China

Supervised Fine-Tuning Reinforcement Learning LLM Fine-Tuning Catastrophic Forgetting

Feb 10, 2026

Reinforcement learning with an expectile-based objective

We consider the policy evaluation and control in a finite horizon reinforcement learning (RL) setting under an expectile-based objective. First, we derive the mean-squared error (MSE) and concentration bounds for the classic estimator of expectiles based on independent and identically distributed (i.i.d.) samples. To the best of our knowledge, expectiles have not been analyzed in the non-asymptotic regime and the bounds we derive may be of independent interest. Next, we analyze a Monte Carlo type estimator of expectile of the Markov chain underlying a given policy. We derive upper bounds that hold in expectation as well as with high probability for this estimator. Further, we show the order-optimality of our estimator by deriving a lower bound on expectile-based policy evaluation. For the problem of control, we adopt a policy gradient approach and derive a policy gradient theorem for expectiles. Using this result, we propose a gradient estimator with a

O\left(1/m\right)

mean-squared error bounds, where

m

is the number of trajectories. Further, under standard assumptions for policy gradient-type algorithms, we establish smoothness of the expectile-sensitive objective, in turn leading to stationary convergence rate bounds for the overall risk-sensitive policy gradient algorithm that we propose. Finally, we conduct numerical experiments to show the utility of expectiles on popular RL benchmarks.

Shrey Rakeshkumar Patel, Sumedh Gupte, Soumen Pachal +2

Indian Institute of Technology Madras · TCS Research, India

Reinforcement Learning Risk-Sensitive RL Policy Gradient Methods

Feb 9, 2026

SoK: The Pitfalls of Deep Reinforcement Learning for Cybersecurity

Deep Reinforcement Learning (DRL) has achieved remarkable success in domains requiring sequential decision-making, motivating its application to cybersecurity problems. However, transitioning DRL from laboratory simulations to bespoke cyber environments can introduce numerous issues. This is further exacerbated by the often adversarial, non-stationary, and partially-observable nature of most cybersecurity tasks. In this paper, we identify and systematize 11 methodological pitfalls that frequently occur in DRL for cybersecurity (DRL4Sec) literature across the stages of environment modeling, agent training, performance evaluation, and system deployment. By analyzing 66 significant DRL4Sec papers (2018-2025), we quantify the prevalence of each pitfall and find an average of over five pitfalls per paper. We demonstrate the practical impact of these pitfalls using controlled experiments in (i) autonomous cyber defense, (ii) adversarial malware creation, and (iii) web security testing environments. Finally, we provide actionable recommendations for each pitfall to support the development of more rigorous and deployable DRL-based security systems.

Shae McFadden, Myles Foley, Elizabeth Bates +5

King’s College London · The Alan Turing Institute · University College London +3

Reinforcement Learning Cybersecurity

Feb 9, 2026

Combining Constrained Sampling and Reinforcement Learning for Manipulation

Training manipulation policies in contact-rich settings is a core challenge in robotics. While Reinforcement Learning (RL) has demonstrated its strength in such settings, it may struggle to sufficiently explore and discover complex manipulation strategies. To address this, we combine two basic ideas: First, designing appropriate reset strategies (the start state distribution of episodes) has shown promise in improving RL exploration and effectiveness. Second, while model-based approaches to finding trajectories through manipulation are hard, recent work showed that model-based approaches to sampling states on constrained manifolds can be highly efficient. Based on these observations, we propose a novel state sampler that enables efficient goal-conditioned RL in complex contact-rich manipulation tasks. Our sampler explicitly takes into account the structure of contact in order to provide a rich covering of diverse contact modes. By combining constrained sampling resets with projected interpolation and curriculum learning, our novel approach outperforms RL without constrained sampling and alternative reset methods, and effectively trains universal, prehensile and non-prehensile manipulation policies in contact-rich whole-body and in-hand settings. See the project page https://www.user.tu-berlin.de/mtoussai/26-CSRL/ for videos and full source code access.

Marc Toussaint, Cornelius V. Braun, Armand Jordana +5

Technische Universität Berlin · Robotics Institute Germany · LAAS-CNRS +1

Reinforcement Learning Non-Prehensile Manipulation Contact-Rich Robotic Manipulation Goal-Conditioned RL

Feb 8, 2026

Horizon Imagination: Efficient On-Policy Rollout in Diffusion World Models

We study diffusion-based world models for reinforcement learning, which offer high generative fidelity but face critical efficiency challenges in control. Current methods either require heavyweight models at inference or rely on highly sequential imagination, both of which impose prohibitive computational costs. We propose Horizon Imagination (HI), an on-policy imagination process for discrete stochastic policies that denoises multiple future observations in parallel. HI incorporates a stabilization mechanism and a novel sampling schedule that decouples the denoising budget from the effective horizon over which denoising is applied while also supporting sub-frame budgets. Experiments on Atari 100K and Craftium show that our approach maintains control performance with a sub-frame budget of half the denoising steps and achieves superior generation quality under varied schedules. Code is available at https://github.com/leor-c/horizon-imagination.

Lior Cohen, Ofir Nabati, Kaixin Wang +2

Technion · Microsoft Research

Reinforcement Learning World Model Learning Diffusion Model Sampling RL for Diffusion Models

Feb 8, 2026

Process Rewards for Outcome-Guided Reasoning Steps in LLM Reinforcement Learning

Mathematical reasoning in large language models has improved substantially with reinforcement learning using verifiable rewards, where final answers can be checked automatically and converted into reliable training signals. Most such pipelines optimize outcome correctness only, which yields sparse feedback for long, multi-step solutions and offers limited guidance on intermediate reasoning errors. Recent work therefore introduces process reward models (PRMs) to score intermediate steps and provide denser supervision. In practice, PRM scores are often imperfectly aligned with final correctness and can reward locally fluent reasoning that still ends in an incorrect answer. When optimized as absolute rewards, such signals can amplify fluent failure modes and induce reward hacking. We propose PROGRS, a framework that leverages PRMs while keeping outcome correctness dominant. PROGRS treats process rewards as relative preferences within outcome groups rather than absolute targets. We introduce outcome-conditioned centering, which shifts PRM scores of incorrect trajectories to have zero mean within each prompt group. It removes systematic bias while preserving informative rankings. PROGRS combines a frozen quantile-regression PRM with a multi-scale coherence evaluator. We integrate the resulting centered process bonus into Group Relative Policy Optimization (GRPO) without auxiliary objectives or additional trainable components. Across MATH-500, AMC, AIME, MinervaMath, and OlympiadBench, PROGRS consistently improves Pass@1 over outcome-only baselines and achieves stronger performance with fewer rollouts. These results show that outcome-conditioned centering enables safe and effective use of process rewards for mathematical reasoning.

Mohammad Rezaei, Jens Lehmann, Sahar Vahdati

Faculty of Computer Science, Technical University of Dresden, Dresden, Germany · Amazon Science · Technical University of Dresden, Dresden, Germany +1

LLM Mathematical Reasoning Reinforcement Learning Process Reward Models Reinforcement Learning with Verifiable Rewards

Feb 8, 2026

USER: A Unified and Extensible System for Real-World Online Policy Learning in Embodied AI

Online policy learning directly in the physical world is a promising yet challenging direction for embodied intelligence. Unlike simulation, real-world systems cannot be arbitrarily accelerated, cheaply reset, or massively replicated, suggesting that real-world policy learning is not merely an algorithmic problem, but inherently a systems problem. We present USER, a Unified and extensible SystEm for real-world online policy leaRning. On the systems side, USER introduces a hardware abstraction layer for unified robot management and an adaptive communication plane that enables efficient cloud-edge training. On the learning side, USER adopts a fully asynchronous training framework, designs a persistent and cache-aware replay buffer, and provides extensible abstractions for rewards, algorithms, and policies. Experiments in both simulation and the real world demonstrate that USER supports multi-robot coordination, heterogeneous manipulators, cloud-edge training with large models, and long-running asynchronous training. Together, these capabilities establish USER as a unified and extensible systems foundation for real-world online policy learning.

Hongzhi Zang, Shu'ang Yu, Hao Lin +14

Tsinghua University · Shanghai AI Laboratory · Infinigence AI +3

Robot Policy Learning Reinforcement Learning Distributed RL Robot Learning

Feb 8, 2026

rePIRL: Learn PRM with Inverse RL for LLM Reasoning

Process rewards have been widely used in deep reinforcement learning to improve training efficiency, reduce variance, and prevent reward hacking. In LLM reasoning, existing works also explore various solutions for learning effective process reward models (PRM) with or without the help of an expert policy. However, existing methods either rely on strong assumptions about the expert policies (e.g., requiring their reward functions) or suffer intrinsic limitations (e.g., entropy collapse), resulting in weak PRMs or limited generalizability. In this paper, we introduce rePIRL, an inverse RL-inspired framework that learns effective PRMs with minimal assumptions about expert policies. Specifically, we design a dual learning process that updates the policy and the PRM interchangeably. Our learning algorithm has customized techniques to address the challenges of scaling traditional inverse RL to LLMs. We theoretically show that our proposed learning framework can unify both online and offline PRM learning methods, justifying that rePIRL can learn PRMs with minimal assumptions. Empirical evaluations on standardized math and coding reasoning datasets demonstrate the effectiveness of rePIRL over existing methods. We further show the application of our trained PRM in test-time training, test-time scaling, and providing an early signal for training hard problems. Finally, we validate our training recipe and key design choices via a detailed ablation study.

Xian Wu, Kaijie Zhu, Ying Zhang +2

Meta AI · Department of Computer Science, University of California, Santa Barbara · Independent Researcher +1

Reinforcement Learning Process Reward Models RL for Language Model Reasoning LLM Reasoning

Feb 7, 2026

A Behavior-first approach to Multi-Intention Inverse Reinforcement Learning

Inverse Reinforcement Learning (IRL) seeks to infer reward functions from expert demonstrations. When demonstrations originate from multiple experts with different intentions, the problem is known as Multi-Intention IRL (MI-IRL). Recent deep generative MI-IRL approaches couple behavior clustering and reward learning, but typically require prior knowledge of the number of true behavioral modes

K^*

. This reliance on expert knowledge limits their adaptability to new behaviors, and only enables analysis related to the learned rewards, and not across the behavior modes used to train them. We propose Contrastive Multi-Intention IRL (CoMI-IRL), a transformer-based unsupervised framework that decouples behavior representation and clustering from downstream reward learning. Our experiments show that CoMI-IRL outperforms existing approaches without a priori knowledge of

K^*

or labels, while allowing for visual interpretation of behavior relationships and adaptation to unseen behavior without full retraining.

Antonio Mone, Frans A. Oliehoek, Luciano Cavalcante Siebert

Delft University of Technology, the Netherlands

Reinforcement Learning

Feb 7, 2026

Scalable Dexterous Robot Learning with AR-based Remote Human-Robot Interactions

This paper focuses on the scalable robot learning for manipulation in the dexterous robot arm-hand systems, where the remote human-robot interactions via augmented reality (AR) are established to collect the expert demonstration data for improving efficiency. In such a system, we present a novel method to address the general manipulation task problem. Specifically, the proposed method consists of two phases: i) In the first phase for pretraining, the policy is created in a behavior cloning (BC) manner, through leveraging the learning data from our AR-based remote human-robot interaction system; ii) In the second phase, a contrastive learning empowered reinforcement learning (RL) method is developed to obtain more efficient and robust policy than the BC, and thus a projection head is designed to accelerate the learning progress. An event-driven augmented reward is adopted for enhancing the safety. To validate the proposed method, both the physics simulations via PyBullet and real-world experiments are carried out. The results demonstrate that compared to the baselines, our method not only significantly speeds up the training process, but also achieves much better performance in terms of the success rate for fulfilling the manipulation tasks. By conducting the ablation study, it is confirmed that the proposed RL with contrastive learning overcomes policy collapse. Supplementary demonstrations are available at https://cyberyyc.github.io/.

Yicheng Yang, Ruijiao Li, Lifeng Wang +7

Reinforcement Learning Robotic RL Robotic Manipulation Robot Imitation Learning Contrastive RL

Feb 6, 2026

F-GRPO: Don't Let Your Policy Learn the Obvious and Forget the Rare

Reinforcement Learning with Verifiable Rewards (RLVR) is commonly based on group sampling to estimate advantages and stabilize policy updates. In practice, computational limits often rule out very large groups, so training proceeds with finite rollout sets that can reinforce only the correct behavior they expose. At practical group sizes, updates can miss rare-correct trajectories while still containing mixed rewards, concentrating probability on more common sampled solutions. We derive the probability of such prompt-local tail-miss events as a function of group size, showing non-monotonic behavior, and in the categorical abstraction characterize how unsampled-correct mass can shrink even as total correct mass grows. Motivated by this analysis, we propose a difficulty-aware scaling coefficient, inspired by Focal loss, that down-weights updates on high-success sampled groups. Empirically, categorical simulation illustrates the same effect in the categorical setting, Maze provides a single-solution test, and LLM experiments include a representative GRPO group-size sweep together with fixed-

N

transfer across GRPO, DAPO, and CISPO. On Qwen2.5-7B at

N{=}8

, our method improves average math pass@256 from 64.1

\rightarrow

70.3 (GRPO), 69.3

\rightarrow

72.5 (DAPO), and 73.2

\rightarrow

76.8 (CISPO); OOD pass@256 also improves in all three cases, without increasing group size or computational cost.

Daniil Plyusov, Alexey Gorbatovski, Boris Shaposhnikov +4

T-Tech · Saint Petersburg Electrotechnical University “LETI”

Reinforcement Learning RL for Language Model Reasoning Group Relative Policy Optimization Reinforcement Learning with Verifiable Rewards

Feb 6, 2026

The impact of temporal resampling on the safety of clinical reinforcement learning: A simulated trial

Reinforcement learning (RL) is a type of artificial intelligence for making optimal choices. In healthcare, researchers generally use offline RL (ORL), where models are trained and evaluated from retrospective observational data. To accommodate naturally irregular clinical records, researchers often resample the data into uniform time intervals before training (known as binning). However, binning may create counterfactual trajectories that do not represent clinical practice. We first quantified these distortions in patients receiving continuous insulin infusions in the MIMIC-IV intensive care database. Binning reversed the apparent causal order for over 70% of paired glucose measurements and insulin adjustments within the same interval, and frequently masked significant dysglycaemic events and merged distinct treatment decisions together. We then conducted an in silico clinical trial using 30 virtual patients with type 1 diabetes from the FDA-approved UVA/Padova simulator. The simulator was modified to include stochastic intervals between decisions and used to generate a training dataset for offline RL. We trained three ORL algorithms on both the unprocessed dataset and equivalent datasets resampled at 10-minute, 2-hour, and 4-hour intervals. When deployed back into the simulated environment, temporal resampling was found to reduce model performance by up to 60% relative to unprocessed data, with 4-hour binning causing all agents to perform worse than the dataset's baseline. Retrospective evaluation on resampled data actively obscured this effect, predicting 1.5-3x better returns than agents achieved in practice. We recommend that future research in this area prioritises datasets with natural clinical timings between decisions, which may be a necessary step before these models can be safely deployed into patient care.

Thomas Frost, Hrisheekesh Vaidya, Steve Harris

University College London, London, United Kingdom · University College London Hospitals NHS Foundation Trust, London, United Kingdom

Reinforcement Learning Offline RL Safe RL

Feb 6, 2026

Difficulty-Estimated Policy Optimization

Recent advancements in Large Reasoning Models (LRMs), exemplified by DeepSeek-R1, have underscored the potential of scaling inference-time compute through Group Relative Policy Optimization (GRPO). However, GRPO frequently suffers from gradient signal attenuation when encountering problems that are either too trivial or overly complex. In these scenarios, the disappearance of inter-group advantages makes the gradient signal susceptible to noise, thereby jeopardizing convergence stability. While variants like DAPO attempt to rectify gradient vanishing, they do not alleviate the substantial computational overhead incurred by exhaustive rollouts on low-utility samples. In this paper, we propose Difficulty-Estimated Policy Optimization (DEPO), a novel framework designed to optimize the efficiency and robustness of reasoning alignment. DEPO integrates an online Difficulty Estimator that dynamically assesses and filters training data before the rollout phase. This mechanism ensures that computational resources are prioritized for samples with high learning potential. Empirical results show that DEPO achieves approximately 1.86x faster rollouts than DAPO in the profiled configuration, while improving benchmark performance over GRPO at comparable training cost. Our approach significantly lowers the computational barrier for training high-performance reasoning models, offering a more sustainable path for reasoning scaling.

Yu Zhao, Fan Jiang, Tianle Liu +5

Alibaba International Digital Commerce · School of Software Technology, Dalian University of Technology Dalian, China

RL for Language Model Reasoning Policy Optimization Reinforcement Learning Group Relative Policy Optimization

Feb 5, 2026

Flow Matching for Offline Reinforcement Learning with Discrete Actions

Generative policies based on diffusion models and flow matching have shown strong promise for offline reinforcement learning (RL), but their applicability remains largely confined to continuous action spaces. To address a broader range of offline RL settings, we extend flow matching to a general framework that supports discrete action spaces with multiple objectives. Specifically, we replace continuous flows with continuous-time Markov chains, trained using a Q-weighted flow matching objective. We then extend our design to multi-agent settings, mitigating the exponential growth of joint action spaces via a factorized conditional path. We theoretically show that, under idealized conditions, optimizing this objective recovers the optimal policy. Extensive experiments further demonstrate that our method performs robustly across diverse settings and benchmarks, including high-dimensional control, multi-agent games, and dynamically changing preferences over multiple objectives, while outperforming traditional offline RL methods in practical multi-modal decision-making scenarios. Our discrete framework can also be applied to continuous-control problems through action quantization, providing a flexible trade-off between representational complexity and performance.

Fairoz Nower Khan, Nabuat Zaman Nahim, Ruiquan Huang +2

Department of Computer Science, University of Kentucky · Department of Computing and Information Sciences, Rochester Institute of Technology

Markov Chains Reinforcement Learning Offline RL Flow Matching Multi-Agent Reinforcement Learning Multi-Objective Reinforcement Learning

Feb 5, 2026

On the Role of Computation in Reinforcement Learning

How does the amount of compute available to a reinforcement learning (RL) policy affect its learning? Can policies using a fixed amount of parameters, still benefit from additional compute? The standard RL framework does not provide a language to answer these questions formally. Empirically, deep RL policies are often parameterized as neural networks with static architectures, conflating the amount of compute and the number of parameters. In this paper, we formalize compute bounded policies and prove that policies which use more compute can solve problems and generalize to longer-horizon tasks that are outside the scope of policies with less compute. Building on prior work in algorithmic learning and model-free planning, we propose a minimal architecture that can use a variable amount of compute. Our experiments complement our theory. On a set 31 different tasks spanning online and offline RL, we show that

(1)

this architecture achieves stronger performance simply by using more compute, and

(2)

stronger generalization on longer-horizon test tasks compared to standard feedforward networks or deep residual network using up to 5 times more parameters.

Raj Ghugare, Michał Bortkiewicz, Alicja Ziarko +1

Department of Computer Science, Princeton University · Warsaw University of Technology · University of Warsaw +2

Reinforcement Learning Model-Free RL Policy Learning

Feb 3, 2026

Chain-of-Goals Hierarchical Policy for Long-Horizon Offline Goal-Conditioned RL

Offline goal-conditioned reinforcement learning remains challenging for long-horizon tasks. While hierarchical approaches mitigate this issue by decomposing tasks, most existing methods rely on separate high- and low-level networks and generate only a single intermediate subgoal, leaving several structural limitations in long-horizon decision-making. To address this limitation, we draw inspiration from chain-of-thought reasoning and propose the Chain-of-Goals Hierarchical Policy (CoGHP), a novel framework that reformulates hierarchical decision-making as autoregressive sequence modeling within a unified architecture. Given a state and a final goal, CoGHP autoregressively generates a sequence of latent subgoals followed by the primitive action, where each latent subgoal acts as a reasoning step that conditions subsequent predictions. To implement this efficiently, we introduce an MLP-Mixer backbone, which supports cross-token communication and captures structural relationships among state, goal, latent subgoals, and action. Across challenging navigation and manipulation benchmarks, CoGHP consistently outperforms strong offline baselines, demonstrating improved performance on long-horizon tasks. Project page: https://wlsdn9350.github.io/projects/coghp/

Jinwoo Choi, Sang-Hyun Lee, Seung-Woo Seo

Department of Electrical and Computer Engineering, Seoul National University, Seoul, South Korea · Department of Automotive Engineering, Ajou University, Gyeonggi-do, South Korea

Reinforcement Learning Hierarchical RL Offline RL Long-Horizon Agent Tasks Goal-Conditioned RL Long-Horizon RL

Feb 2, 2026

VLM-Guided Experience Replay

Recent advances in Large Language Models (LLMs) and Vision-Language Models (VLMs) have enabled powerful semantic and multimodal reasoning capabilities, creating new opportunities to enhance sample efficiency, high-level planning, and interpretability in reinforcement learning (RL). While prior work has integrated LLMs and VLMs into various components of RL, the replay buffer, a core component for storing and reusing experiences, remains unexplored. We propose addressing this gap by leveraging VLMs to guide the prioritization of experiences in the replay buffer. Our key idea is to use a frozen, pre-trained VLM as an automated evaluator to identify and prioritize promising sub-trajectories from the agent's experiences. Across scenarios, including game-playing and robotics, spanning both discrete and continuous domains, agents trained with our proposed prioritization method achieve 15-57% higher average success rates and improve sample efficiency by 35-55% compared to previous approaches. Project page: https://esharony.me/projects/vlm-rb/

Elad Sharony, Tom Jurgenson, Orr Krupnik +2

Technion · ForSight Robotics · Nvidia Research

Reinforcement Learning Prioritized Experience Replay Sample-Efficient RL

Feb 2, 2026

Data- and Variance-dependent Regret Bounds for Online Tabular MDPs

This work studies online episodic tabular Markov decision processes (MDPs) with known transitions and develops best-of-both-worlds algorithms that achieve refined data-dependent regret bounds in the adversarial regime and variance-dependent regret bounds in the stochastic regime. We quantify MDP complexity using a first-order quantity and several new data-dependent measures for the adversarial regime, including a second-order quantity and a path-length measure, as well as variance-based measures for the stochastic regime. To adapt to these measures, we develop algorithms based on global optimization and policy optimization, both built on optimistic follow-the-regularized-leader with log-barrier regularization. For global optimization, our algorithms achieve first-order, second-order, and path-length regret bounds in the adversarial regime, and in the stochastic regime, they achieve a variance-aware gap-independent bound and a variance-aware gap-dependent bound that is polylogarithmic in the number of episodes. For policy optimization, our algorithms achieve the same data- and variance-dependent adaptivity, up to a factor of the episode horizon, by exploiting a new optimistic

Q

-function estimator. Finally, we establish regret lower bounds in terms of data-dependent complexity measures for the adversarial regime and a variance measure for the stochastic regime, implying that the regret upper bounds achieved by the global-optimization approach are nearly optimal.

Mingyi Li, Taira Tsuchiya, Kenji Yamanishi

Department of Mathematical Informatics, The University of Tokyo, Tokyo, Japan · RIKEN AIP, Tokyo, Japan

Regret Minimization in RL Markov Decision Processes Reinforcement Learning Policy Optimization Follow-the-Regularized-Leader

Jan 30, 2026

MASC: Metal-Aware Sampling and Correction via Reinforcement Learning for Accelerated MRI

Metal implants in MRI cause severe artifacts that degrade image quality and hinder clinical diagnosis. Traditional approaches address metal artifact reduction (MAR) and accelerated MRI acquisition as separate problems. We propose MASC, a unified reinforcement learning framework that jointly optimizes metal-aware k-space sampling and artifact correction for accelerated MRI. To enable supervised training, we construct a paired MRI dataset using physics-based simulation, generating k-space data and reconstructions for phantoms with and without metal implants. This paired dataset provides simulated 3D MRI scans with and without metal implants, where each metal-corrupted sample has an exactly matched clean reference, enabling direct supervision for both artifact reduction and acquisition policy learning. We formulate active MRI acquisition as a sequential decision-making problem, where an artifact-aware Proximal Policy Optimization (PPO) agent learns to select k-space phase-encoding lines under a limited acquisition budget. The agent operates on undersampled reconstructions processed through a U-Net-based MAR network, learning patterns that maximize reconstruction quality. We further propose an end-to-end training scheme where the acquisition policy learns to select k-space lines that best support artifact removal while the MAR network simultaneously adapts to the resulting undersampling patterns. Experiments demonstrate that MASC's learned policies outperform conventional sampling strategies, and end-to-end training improves performance compared to using a frozen pre-trained MAR network, validating the benefit of joint optimization. Cross-dataset experiments on FastMRI with physics-based artifact simulation further confirm generalization to realistic clinical MRI data. The code and models of MASC have been made publicly available: https://github.com/hrlblab/masc

Zhengyi Lu, Ming Lu, Chongyu Qu +11

Vanderbilt University, Nashville, TN, USA 37215 · Vanderbilt University Medical Center, Nashville, TN, USA 37232 · Oak Ridge National Laboratory, Oak Ridge, TN, USA 37831

Magnetic Resonance Imaging Accelerated MRI Reconstruction Image Enhancement Reinforcement Learning

Jan 30, 2026

Offline Reinforcement Learning of High-Quality Behaviors Under Robust Style Alignment

We study offline reinforcement learning of style-conditioned policies using explicit style supervision via subtrajectory labeling functions. In this setting, aligning style with high task performance is particularly challenging due to distribution shift and inherent conflicts between style and reward. Existing methods, despite introducing numerous definitions of style, often fail to reconcile these objectives effectively. To address these challenges, we propose a unified definition of behavior style and instantiate it into a practical framework. Building on this, we introduce Style-Conditioned Implicit Q-Learning (SCIQL), which leverages offline goal-conditioned RL techniques, such as hindsight relabeling and value learning, and combine it with a new Gated Advantage Weighted Regression mechanism to efficiently optimize task performance while preserving style alignment. Experiments demonstrate that SCIQL achieves superior performance on both objectives compared to prior offline methods. Code, datasets and visuals are available in: https://mathieu-petitbois.github.io/projects/sciql/.

Mathieu Petitbois, Rémy Portelas, Sylvain Lamprier

Ubisoft La Forge, Bordeaux, France · Université d’Angers, Angers, France

Reinforcement Learning Offline RL

Jan 30, 2026

Emergence of Physical Intelligence via Controllable Information Production

Intrinsic Motivation (IM) aims to train agents without external rewards, enabling useful behavior to emerge from the agent's interaction with its environment alone. However, the dominant IM approaches rely on information-theoretic quantities with designer-chosen variables, introducing bias and lacking a principled connection to dynamics or optimal control (OC). We introduce Controllable Information Production (CIP), a novel IM method explicitly grounded in dynamical systems and OC. CIP measures the rate at which an agent produces information, capturing controllable complexity without external knowledge or bias. CIP unifies IM and OC into a single framework, formalizing physical intelligence as the control of information production. It further reveals connections between the structure of the value function and Kolmogorov-Sinai entropy. CIP consistently outperforms prior IM methods on standard benchmarks in robot learning and solves tasks they fail on, including humanoid self-righting. These results support an organizing principle: that physical intelligence emerges from driving systems toward the edge of controllable chaos.

Tristan Shah, Stas Tiomkin

Department of Computer Science, Texas Tech University

Robotic RL Reinforcement Learning Optimal Control

Jan 29, 2026

Expected Return Causes Outcome-Level Mode Collapse in Reinforcement Learning and How to Fix It with Inverse Probability Scaling

Many reinforcement learning (RL) problems admit multiple terminal solutions of comparable quality, where the goal is not to identify a single optimum but to represent a diverse set of high-quality outcomes. Nevertheless, policies trained by standard expected return maximization routinely collapse onto a small subset of outcomes, a phenomenon commonly attributed to insufficient exploration or weak regularization. We show that this explanation is incomplete: outcome level mode collapse is a structural consequence of the expected-return objective itself. Under idealized learning dynamics, the log-probability ratio between any two outcomes evolves linearly in their reward difference, implying exponential ratio divergence and inevitable collapse independent of the exploration strategy, entropy regularization, or optimization algorithm. We identify the source of this pathology as the probability multiplier inside the expectation and propose a minimal correction: inverse probability scaling, which removes outcome-frequency amplification from the learning signal, fundamentally changes the learning dynamics, and provably yields reward-proportional terminal distributions, preventing collapse in multimodal settings. We instantiate this principle in Group Relative Policy Optimization (GRPO) as a drop-in modification, IPS-GRPO, requiring no auxiliary models or architectural changes. Across different reasoning and molecular generation tasks, IPS-GRPO consistently reduces outcome-level mode collapse while matching or exceeding baseline performance, suggesting that correcting the objective rather than adding exploration heuristics is key to reliable multimodal policy optimization.

Abhijeet Sinha, Sundari Elango, Dianbo Liu

National University of Singapore, Singapore · IIT Madras, India

Reinforcement Learning Behavioral Diversity in RL Group Relative Policy Optimization

Jan 26, 2026

Just-In-Time Reinforcement Learning: Continual Learning in LLM Agents Without Gradient Updates

While Large Language Model (LLM) agents excel at general tasks, they inherently struggle with continual adaptation due to the frozen weights after deployment. Conventional reinforcement learning (RL) offers a solution but incurs prohibitive computational costs and the risk of catastrophic forgetting. We introduce Just-In-Time Reinforcement Learning (JitRL), a training-free framework that enables test-time policy optimization without any gradient updates. JitRL maintains a dynamic, non-parametric memory of experiences and retrieves relevant trajectories to estimate action advantages on-the-fly. These estimates are then used to directly modulate the LLM's output logits. We theoretically prove that this additive update rule is the exact closed-form solution to the KL-constrained policy optimization objective. Extensive experiments on WebArena and Jericho demonstrate that JitRL establishes a new state-of-the-art among training-free methods. Crucially, JitRL outperforms the performance of computationally expensive fine-tuning methods (e.g., WebRL) while reducing monetary costs by over 30 times, offering a scalable path for continual learning agents. The code is available at https://github.com/liushiliushi/JitRL.

Yibo Li, Zijie Lin, Ailin Deng +5

National University of Singapore, Singapore.

Continual Learning for LLM Agents Reinforcement Learning Test-Time RL Policy Learning RL for LLM Agents Continual RL+1

Jan 21, 2026

CLEANER: Self-Purified Trajectories Boost Agentic Reinforcement Learning

Agentic Reinforcement Learning (RL) has empowered Large Language Models (LLMs) to utilize tools like Python interpreters for complex problem-solving. However, for parameter-constrained models (e.g., 4B--7B), the exploration phase is often plagued by frequent execution failures, creating noisy trajectories that hinder policy optimization. Under standard outcome-based reward settings, this noise leads to a critical credit assignment issue, where erroneous actions are inadvertently reinforced alongside successful outcomes. Existing mitigations face a dilemma: dense rewards often trigger reward hacking, while supersampling incurs prohibitive computational costs. To address these challenges, we propose CLEANER. Distinct from external filtering methods, CLEANER exploits the model's intrinsic self-correction capabilities to eliminate error-contaminated context directly during data collection. At its core, the Similarity-Aware Adaptive Rollback (SAAR) mechanism autonomously constructs clean, purified trajectories by retrospectively replacing failures with successful self-corrections. Based on semantic similarity, SAAR adaptively regulates replacement granularity from shallow execution repairs to deep reasoning substitutions. By training on these self-purified paths, the model internalizes correct reasoning patterns rather than error-recovery loops. Empirical results on AIME24/25, GPQA, and LiveCodeBench show average accuracy gains of 6%, 3%, and 5% over baselines. Notably, CLEANER matches state-of-the-art performance using only one-third of the training steps, highlighting trajectory purification as a scalable solution for efficient agentic RL. Our models and code are available at GitHub

Tianshi Xu, Yuteng Chen, Meng Li

Peking University · NTU, Singapore

Agentic RL Reinforcement Learning RL for Language Model Reasoning Sample-Efficient RL RL for LLM Agents

Jan 21, 2026

Reinforcement Learning in the Real World: A Survey of Statistical Challenges and Future Directions

Reinforcement learning (RL) has achieved remarkable success in real-world decision-making across diverse domains, including gaming, robotics, online advertising, public health, and natural language processing. Despite these advances, a substantial gap remains between RL research and its deployment in many practical settings. Two recurring challenges often underlie this gap. First, many settings offer limited opportunity for the agent to interact extensively with the target environment due to practical constraints. Second, many target environments often undergo substantial changes, requiring redesign and redeployment of RL systems (e.g., advancements in science and technology that change the landscape of healthcare delivery). Addressing these challenges and bridging the gap between basic research and application requires theory and methodology that directly inform the design, implementation, and continual improvement of RL systems in real-world settings. In this paper, we frame the application of RL in practice as a three-component process: (i) online learning and optimization during deployment, (ii) post- or between-deployment offline analyses, and (iii) repeated cycles of deployment and redeployment to continually improve the RL system. We provide a narrative review of recent advances that address the statistical challenges arising across these three components, including methods for enhancing sample efficiency during online deployment, maximizing data utility for post- or between-deployment inference, and designing sequences of deployments for continual improvement. We also outline future research directions in RL that are use-inspired -- aiming for impactful application of RL in practice.

Asim H. Gazi, Yongyi Guo, Daiqi Gao +3

Department of Computer Science, Harvard University · Department of Statistics, Harvard University · Department of Statistics, University of Wisconsin–Madison +2

Non-Stationary RL Reinforcement Learning Sample-Efficient RL