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

1,767 papers195 papers in the last four weeks, down 8% on the four weeks before. 1.3% of all new papers.

May 6, 2026

Reinforcement Learning for Compositional Generalization with Outcome-Level Optimization

Compositional generalization refers to correctly interpret novel combinations of known primitives, which remains a major challenge. Existing approaches often rely on supervised fine-tuning, which encourages models to imitate target outputs. This token-level training paradigm fails to capture the global compositional structure required for generalizing to unseen combinations. In this work, we investigate whether compositional generalization can instead be improved through outcome-level reinforcement learning. We adopt Group Relative Policy Optimization to optimize models based on feedback on their final outputs. Within this framework, we explore both a simple binary outcome reward and a composite reward that provides additional composition feedback. Experiments on multiple compositional benchmarks show that reinforcement learning improves compositional generalization compared to supervised fine-tuning. Further analysis reveals that supervised models tend to overfit frequent training compositions, whereas reinforcement learning improves compositional generalization by reshaping the output distribution, particularly for more complex composition types.

Xiyan Fu, Wei Liu

Nanyang Technological University · Independent Researcher

Reinforcement Learning RL for Generative Models RL Fine-Tuning Compositional Generalization

May 6, 2026

A Harmonic Mean Formulation of Average Reward Reinforcement Learning in SMDPs

Recent research has revived and amplified interest in algorithms for undiscounted average reward reinforcement learning in infinite-horizon, non-episodic (continuing) tasks. Semi-Markov decision processes (SMDPs) are of particular interest. In SMDPs, discrete actions stochastically generate both rewards and durations, and the objective is to optimize the average reward rate. Existing algorithms approach this by optimizing the ratio of rewards to durations. However, when rewards and durations are non-stationary (in the infinite horizon), this can be incorrect. This paper presents a novel modified harmonic mean operator that correctly computes reward rates even under such conditions. This yields model-free learning algorithms that can work with SMDPs, while maintaining robustness to non-stationary reward and duration distributions over time. We prove theoretical properties of the modified harmonic mean operator, and empirically demonstrate its efficacy in comparison to existing algorithms.

Erel Shtossel, Alicia Vidler, Uri Shaham +1

Bar Ilan University Ramat Gan, Israel

Non-Stationary RL Reinforcement Learning Average-Reward RL Sequential Decision Making Semi-Markov Decision Process Model-Free RL

May 6, 2026

SPHERE: Mitigating the Loss of Spectral Plasticity in Mixture-of-Experts for Deep Reinforcement Learning

In deep reinforcement learning (DRL), an agent is trained from a stream of experience. In a continual learning setting, such agents can suffer from plasticity loss: their ability to learn new skills from new experiences diminishes over training. Recently, Mixture-of-Experts (MoE) networks have been reported to enable scaling laws and facilitate the learning of diverse skills. However, in continual reinforcement learning settings, their performance can degenerate as learning proceeds, indicating a loss of plasticity. To address this, building on Neural Tangent Kernel (NTK) theory, we formalize the plasticity loss in MoE policies as a loss of spectral plasticity. We then derive a tractable proxy for spectral plasticity, one expressible in terms of individual expert feature matrices. Leveraging this proxy, we introduce SPHERE, a practical Parseval penalty tailored for MoE-based policies that alleviates the loss of spectral plasticity. On MetaWorld and HumanoidBench, SPHERE improves average success under continual RL by 133% and 50% over an unregularized MoE baseline, while maintaining higher spectral plasticity throughout training.

Lirui Luo, Guoxi Zhang, Hongming Xu +2

State Key Lab of General AI, School of Intelligence Science and Technology, Peking University · State Key Laboratory of General Artificial Intelligence, BIGAI · Institute for Artificial Intelligence, Peking University

Loss of Plasticity Reinforcement Learning Mixture of Experts Neural Network Training Dynamics Policy Learning Continual RL

May 6, 2026

Power Distribution Bridges Sampling, Self-Reward RL, and Self-Distillation

Recent analyses question whether reinforcement learning (RL) is responsible for strong reasoning in large language models (LLMs). At the same time, distillation and inference-time sampling, including power sampling, have emerged as effective ways to improve LLM performance. However, the relationship among RL, distillation, and sampling remains unclear. In this study, we focus on the power distribution, the target distribution of power sampling, and show that the power distribution bridges sampling, self-reward KL-regularized RL, and self-distillation. From the sampling perspective, we show that inexpensive local approximations cannot reproduce sequence-level power without information about possible suffixes. From the RL perspective, the power distribution is the closed-form optimizer of KL-regularized RL when the model's sequence-level log-probabilities are used as the reward. This identification leads to power self-distillation, an offline distillation surrogate that shares the same target distribution and amortizes the cost of power sampling into supervised training on teacher samples. We further show that power self-distillation can achieve self-reward sharpening, that downstream true-reward gains are governed by the covariance between true reward and self-reward under the power distribution, and that this ideal power-teacher gain transfers to a distilled student when the student is close to the power teacher. Experiments on reasoning tasks support our analysis: power sampling raises self-reward, true-reward gains depend on alignment with self-reward, and power self-distillation can match or exceed the performance of power sampling at much lower inference cost. The code is available at https://github.com/tom4649/power-self-distillation.

Akiyoshi Tomihari, Issei Sato

Department of Computer Science, The University of Tokyo

Reinforcement Learning KL-Regularized RL RL for Language Model Reasoning Language Model Distillation Self-Distillation

May 6, 2026

CRAFT: Counterfactual-to-Interactive Reinforcement Fine-Tuning for Driving Policies

Open-loop imitation learning has advanced modern autonomous driving policy architectures, but closed-loop deployment remains vulnerable to policy-induced distribution shift. Existing post-training paradigms exhibit fundamental trade-offs: closed-loop RL fine-tuning provides grounded feedback from executed actions but is constrained by the sparsity of informative events, whereas counterfactual fine-tuning provides dense supervision over candidate futures but inherits bias from imperfect future estimates. We introduce Counterfactual-to-Interactive Reinforcement Fine-Tuning (CRAFT), an on-policy framework that formulates closed-loop post-training as proxy-residual optimization. CRAFT uses group-normalized counterfactual advantages as a dense proxy for real closed-loop advantages and aligns this proxy with the closed-loop world through grounded residual correction from interaction-critical events. To stabilize adaptation, CRAFT regularizes the online policy toward an EMA teacher via asymmetric KL self-distillation. Theoretically, CRAFT decomposes the real closed-loop policy gradient into proxy and residual terms under the same visited-state distribution, reducing residual variance with an aligned proxy while mitigating proxy bias through grounded residual approximation. Empirically, CRAFT achieves the strongest closed-loop gains on Bench2Drive across hierarchical planning, vision-language-action, and vocabulary-scoring architectures. Ablations, scaling behavior, stability analyses, and transfer results further validate the complementary roles of dense counterfactual proxy and grounded residual correction. Project page: https://currychen77.github.io/CRAFT.

Keyu Chen, Nanfei Ye, Yida Wang +4

School of Vehicle and Mobility, Tsinghua University · Li Auto Inc

Reinforcement Learning Autonomous Driving Planning RL Post-Training RL Fine-Tuning Counterfactual Learning

May 6, 2026

Towards Robust LLM Post-Training: Automatic Failure Management for Reinforcement Fine-Tuning

Reinforcement fine-tuning (RFT) has become a core paradigm for post-training large language models, yet its training process remains highly fragile. Existing efforts mainly improve reliability at the system level or address specific issues in individual subproblems by modifying RFT algorithms. Despite their effectiveness, they largely overlook the problem of failure management at the training-process level. When training goes wrong, practitioners still rely heavily on expert-driven manual inspection and correction, and automatic failure management for RFT remains largely unexplored. In this paper, we take a first step toward systematic failure management for reinforcement fine-tuning. To understand the empirical structure of RFT failures, we first construct RFT-FaultBench, the first benchmark for fine-grained failures in reinforcement fine-tuning, covering 5 fault families, 16 fault types, 779 training runs, 22,549 train-step records, and 1,457,288 trajectory-level records. Based on this benchmark, we conduct a comprehensive empirical study showing that RFT failures are both observable from training dynamics and distinguishable through their empirical fault fingerprints. Building on these findings, we propose RFT-FM, an automatic failure management framework for reinforcement fine-tuning that unifies anomaly detection, failure diagnosis, and auto remediation in a closed loop. Experimental results show that RFT-FaultBench is neither trivial nor saturated: it exhibits clear anomaly structure while still posing substantial challenges, especially under subtle fault settings. Moreover, RFT-FM shows strong capability in detecting, diagnosing, and mitigating RFT failures.

Lingzhe Zhang, Tong Jia, Yunpeng Zhai +6

Peking University, Beijing, China · Alibaba Group, China · University of Illinois Chicago, United States

Reinforcement Learning RL Fine-Tuning

May 6, 2026

Extending Differential Temporal Difference Methods for Episodic Problems

Differential temporal difference (TD) methods are value-based reinforcement learning algorithms that have been proposed for infinite-horizon problems. They rely on reward centering, where each reward is centered by the average reward. This keeps the return bounded and removes a value function's state-independent offset. However, reward centering can alter the optimal policy in episodic problems, limiting its applicability. Motivated by recent works that emphasize the role of normalization in streaming deep reinforcement learning, we study reward centering in episodic problems and propose a generalization of differential TD. We prove that this generalization maintains the ordering of policies in the presence of termination, and thus extends differential TD to episodic problems. We show equivalence with a form of linear TD, thereby inheriting theoretical guarantees that have been shown for those algorithms. We then extend several streaming reinforcement learning algorithms to their differential counterparts. Across a range of base algorithms and environments, we empirically validate that reward centering can improve sample efficiency in episodic problems.

Kris De Asis, Mohamed Elsayed, Jiamin He

1Openmind Research Institute · Department of Computing Science, University of Alberta, Canada · 3Alberta Machine Intelligence Institute (Amii)

Episodic RL Reinforcement Learning Temporal-Difference Learning

May 5, 2026

Hierarchical Support Vector State Partitioning for Distilling Black Box Reinforcement Learning Policies

We introduce State Vector Space Partitioning (SVSP), a novel method to mimic a black box reinforcement learning policy using a set of human-interpretable subpolicies. By partitioning a distillation dataset of state action pairs with linear support vector machine splits, SVSP constructs a compact and structured representation of the original policy. Our method improves mean return by +7.4% over previous critic driven state partitioning attempts such as Voronoi State Partitioning (VSP) and +2.8% over the original TD3 policy, while reducing the number of required subpolicies against VSP by 82.1%. Our results pave the path towards a more flexible form of distillation where both the decision boundary and surrogate models can be chosen within a margin of the original black box behavior.

Senne Deproost, Mehrdad Asadi, Ann Nowé

aVrije Universiteit Brussel · bFlanders Make

Reinforcement Learning Interpretability in RL Policy Learning

May 5, 2026

Optimal Posterior Sampling for Policy Identification in Tabular Markov Decision Processes

We study the

(\varepsilon, δ)

-PAC policy identification problem in finite-horizon episodic Markov Decision Processes. Existing approaches provide finite-time guarantees for approximate settings (

\varepsilon>0

) but suffer from high computational cost, rendering them hard to implement, and also suffer from suboptimal dependence on

\log(1/δ)

. We propose a randomized and computationally efficient algorithm for best policy identification that combines posterior sampling with an online learning algorithm to guide exploration in the MDP. Our method achieves asymptotic optimality in sample complexity, also in terms of posterior contraction rate, and runs in

O(S^2AH)

per episode, matching standard model-based approaches. Unlike prior algorithms such as MOCA and PEDEL, our guarantees remain meaningful in the asymptotic regime and avoid sub-optimal polynomial dependence on

\log(1/δ)

. Our results provide both theoretical insights and practical tools for efficient policy identification in tabular MDPs.

Cyrille Kone, Kevin Jamieson

Univ. Lille, CNRS, Inria, Centrale Lille, UMR 9189-CRIStAL, F-59000 Lille, France · University of Washington, Seattle, USA

Thompson Sampling Markov Decision Processes Reinforcement Learning Sample-Efficient RL

May 5, 2026

Discovering Reinforcement Learning Interfaces with Large Language Models

Reinforcement learning systems rely on environment interfaces that specify observations and reward functions, yet constructing these interfaces for new tasks often requires substantial manual effort. While recent work has automated reward design using large language models (LLMs), these approaches assume fixed observations and do not address the broader challenge of synthesizing complete task interfaces. We study RL task interface discovery from raw simulator state, where both observation mappings and reward functions must be generated. We propose LIMEN (Code available at https://github.com/Lossfunk/LIMEN), a LLM guided evolutionary framework that produces candidate interfaces as executable programs and iteratively refines them using policy training feedback. Across novel discrete gridworld tasks and continuous control domains spanning locomotion and manipulation, joint evolution of observations and rewards discovers effective interfaces given only a trajectory-level success metric, while optimizing either component alone fails on at least one domain. These results demonstrate that automatic construction of RL interfaces from raw state can substantially reduce manual engineering and that observation and reward components often benefit from co-design, as single-component optimization fails catastrophically on at least one domain in our evaluation suite.

Akshat Singh Jaswal, Ashish Baghel, Paras Chopra

Lossfunk

Reinforcement Learning LLM-Guided RL Reward Design for RL Simulation-Based RL

May 4, 2026

Investigating Action Encodings in Recurrent Neural Networks in Reinforcement Learning

Building and maintaining state to learn policies and value functions is critical for deploying reinforcement learning (RL) agents in the real world. Recurrent neural networks (RNNs) have become a key point of interest for the state-building problem, and several large-scale reinforcement learning agents incorporate recurrent networks. While RNNs have become a mainstay in many RL applications, many key design choices and implementation details responsible for performance improvements are often not reported. In this work, we discuss one axis on which RNN architectures can be (and have been) modified for use in RL. Specifically, we look at how action information can be incorporated into the state update function of a recurrent cell. We discuss several choices in using action information and empirically evaluate the resulting architectures on a set of illustrative domains. Finally, we discuss future work in developing recurrent cells and discuss challenges specific to the RL setting.

Matthew Schlegel, Volodymyr Tkachuk, Adam White +1

University of Alberta

Reinforcement Learning Recurrent Neural Networks

May 4, 2026

Enhancing RL Generalizability in Robotics through SHAP Analysis of Algorithms and Hyperparameters

Despite significant advances in Reinforcement Learning (RL), model performance remains highly sensitive to algorithm and hyperparameter configurations, while generalization gaps across environments complicate real-world deployment. Although prior work has studied RL generalization, the relative contribution of specific configurations to the generalization gap has not been quantitatively decomposed and systematically leveraged for configuration selection. To address this limitation, we propose an explainable framework that evaluates RL performance across robotic environments using SHapley Additive exPlanations (SHAP) to quantify configuration impacts. We establish a theoretical foundation connecting Shapley values to generalizability, empirically analyze configuration impact patterns, and introduce SHAP-guided configuration selection to enhance generalization. Our results reveal distinct patterns across algorithms and hyperparameters, with consistent configuration impacts across diverse tasks and environments. By applying these insights to configuration selection, we achieve improved RL generalizability and provide actionable guidance for practitioners.

Lingxiao Kong, Cong Yang, Oya Deniz Beyan +1

Fraunhofer Institute for Applied Information Technology FIT, Germany · Soochow University, China · University of Cologne, Germany +1

RL for Robotics Reinforcement Learning Robot Policy Generalization Interpretability in RL RL HPO

May 4, 2026

Recurrent Deep Reinforcement Learning for Chemotherapy Control under Partial Observability

Chemotherapy dose optimization can be formulated as a dynamic treatment regime, requiring sequential decisions under uncertainty that must balance tumor suppression against toxicity. However, most reinforcement learning approaches assume full observability of the patient state, a condition rarely met in clinical practice. We investigate whether memory-augmented policies can improve chemotherapy control under partial observability. To this end, we employ a recurrent TD3-based approach with separate LSTM actor-critic networks and evaluate it on the AhnChemoEnv benchmark from DTR-Bench, considering both off-policy and on-policy recurrent architectures against feed-forward TD3 and Soft Actor-Critic. Pharmacokinetic and pharmacodynamic variability are held fixed to isolate hidden-state uncertainty and observation noise and to avoid confounding effects from inter-patient variability. Across ten random seeds, recurrence yields modest benefit under full observability but substantially stronger and more stable performance under partial observability, with more consistent tumor suppression and improved normal-cell preservation. These findings indicate that memory-based policies are particularly beneficial when clinically relevant state information is incomplete or noisy.

Firas Mohamed Elamine Kiram, Imane Youkana, Rachida Saouli +2

Dept. of Information Engineering University of Padova Padova, Italy · LINFI Lab, Dept. of Computer Science University of Biskra Biskra, Algeria

Reinforcement Learning Healthcare Sequential Decision Making Partially Observable RL Oncology

May 4, 2026

Reference-Sampled Boltzmann Projection for KL-Regularized RLVR: Target-Matched Weighted SFT, Finite One-Shot Gaps, and Policy Mirror Descent

Online reinforcement learning with verifiable rewards (RLVR) turns checkable outcomes into a scalable training signal, but it keeps rollout generation, verifier scoring, and reference-policy evaluations on the optimization path. Static weighted supervised fine-tuning (SFT) on precomputed rollouts seems to remove this bottleneck, yet a weighted likelihood is not specified by rewards alone: its sampler and weights induce the policy being fit. This paper identifies the reference-sampled weighted-SFT objective whose induced policy equals the fixed-reference KL-regularized RLVR optimizer. The optimizer is the standard Boltzmann target policy, obtained by exponentially tilting the reference policy by verifier reward. Matching a weighted-SFT induced policy to this target forces density-ratio weights; in the reference-sampled subclass, this reduces uniquely, up to prompt scaling, to the prompt-normalized Boltzmann weight

\exp(r(x,y)/β)/Z(x)

. BOLT, a Boltzmann-Targeted SFT procedure, is the empirical estimator of this projection. The finite one-shot analysis separates the exact stored-support price

β\log(1/π^*(S_N\mid x))

from partition estimation, effective-sample-size variance, generalization, optimization, and approximation errors. This decomposition explains why extra SFT epochs cannot repair missing reference-policy coverage and exposes the temperature--coverage--variance frontier. When coverage needs adaptive sampling, refreshed Boltzmann projections become KL policy mirror descent; finite inner solves enter as additive drift from the exact mirror step. Single-run Qwen experiments provide projection evidence for the target-matched weight, one-shot saturation, refreshed-sampler gains, and optimization-time savings, within the stated single-run scope.

Yao Shu, Chenxing Wei, Hongbin Lin +2

Supervised Fine-Tuning Reinforcement Learning KL-Regularized RL Policy Optimization Reinforcement Learning with Verifiable Rewards

May 4, 2026

Binary Rewards and Reinforcement Learning: Fundamental Challenges

Reinforcement learning with verifiable rewards (RLVR) has become a standard approach for improving reasoning in language models, yet models trained with RLVR often suffer from diversity collapse: while single-sample accuracy improves, multi-sample coverage degrades, sometimes falling below the base model. We provide a structural account of this phenomenon grounded in the properties of binary rewards. Binary rewards create a fundamental degeneracy for policy gradient methods: the set of distributions maximizing expected reward is infinite, with no distinguished element. KL-control resolves this degeneracy by selecting, in the limit

β\to 0

, the filtered model

p_*:=a(\cdot\mid\mathcal{Y}_1)

-- the base model conditioned on validity -- which is the unique fully valid distribution closest to the base model in KL divergence. This selection operates through a nontrivial asymmetry: the tilted distribution

p_{[β]}\propto a(y)\,e^{v(y)/β}

converges to

p_*

in forward KL as

β\to 0

, yet

p_*

cannot serve as a direct optimization target because

\mathrm{KL}(q\,\|\,p_*)

is infinite for any full-support policy

q

. We develop explicit formulas relating the hyperparameter

β

to the more interpretable target validity rate

μ

. Under model misspecification -- the typical practical regime -- the pressure to decrease

β

drives the optimizer toward highly concentrated distributions over a small number of valid outputs, collapsing toward ever fewer as

β

decreases, rather than toward the filtered model. We illustrate this mechanism on a toy autoregressive experiment and discuss how alternative divergences that target

p_*

directly -- as pursued empirically by \citet{kruszewski_whatever_2026} -- avoid this failure mode by rewarding coverage of

p_*

's support rather than concentration on high-validity outputs.

Marc Dymetman

∗Scientific Consultant

Reinforcement Learning KL-Regularized RL RL for Language Model Reasoning Reinforcement Learning with Verifiable Rewards Reward Design for RL Behavioral Diversity in RL

May 4, 2026

T $^2$ PO: Uncertainty-Guided Exploration Control for Stable Multi-Turn Agentic Reinforcement Learning

Recent progress in multi-turn reinforcement learning (RL) has significantly improved reasoning LLMs' performances on complex interactive tasks. Despite advances in stabilization techniques such as fine-grained credit assignment and trajectory filtering, instability remains pervasive and often leads to training collapse. We argue that this instability stems from inefficient exploration in multi-turn settings, where policies continue to generate low-information actions that neither reduce uncertainty nor advance task progress. To address this issue, we propose Token- and Turn-level Policy Optimization (T

^2

PO), an uncertainty-aware framework that explicitly controls exploration at fine-grained levels. At the token level, T

^2

PO monitors uncertainty dynamics and triggers a thinking intervention once the marginal uncertainty change falls below a threshold. At the turn level, T

^2

PO identifies interactions with negligible exploration progress and dynamically resamples such turns to avoid wasted rollouts. We evaluate T

^2

PO in diverse environments, including WebShop, ALFWorld, and Search QA, demonstrating substantial gains in training stability and performance improvements with better exploration efficiency. Code is available at: https://github.com/WillDreamer/T2PO.

Haixin Wang, Hejie Cui, Chenwei Zhang +7

University of California, Los Angeles · 2Amazon.com Inc.

RL for LLM Agents Reinforcement Learning Agentic RL Uncertainty-Aware Exploration RL Exploration Multi-Turn RL+2

May 4, 2026

Planner Matters! An Efficient and Unbalanced Multi-agent Collaboration Framework for Long-horizon Planning

Language model (LM)-based agents have demonstrated promising capabilities in automating complex tasks from natural language instructions, yet they continue to struggle with long-horizon planning and reasoning. To address this, we propose an enhanced multi-agent framework that decomposes automation into three roles: a planner for high-level decision-making, an actor for task execution, and a memory manager for contextual reasoning. While this modular decomposition aligns with established design patterns, our core contribution lies in a systematic compute-allocation analysis, revealing that planning is the dominant factor influencing task performance. Execution and memory management require significantly less compute and model capacity to achieve competitive results. Building on these insights, we introduce a planner-centric reinforcement learning approach, which exclusively optimizes the planner using trajectory-level rewards from a VLM-as-judge, while freezing the other components. Extensive experiments on benchmarks spanning web navigation, OS control, and tool use demonstrate that concentrating model capacity and learning on high-level planning yields robust and compute-efficient improvements in long-horizon agent automation. Our code is publicly released.

Wenyi Wu, Sibo Zhu, Kun Zhou +1

University of California, San Diego

Reinforcement Learning Multi-Agent Collaboration Long-Horizon LLM Agents LLM Agent Planning RL for LLM Agents

May 4, 2026

Combining Trained Models in Reinforcement Learning

Deep reinforcement learning (DRL) has delivered strong results in domains such as Atari and Go, but it still suffers from high sample cost and weak transfer beyond the training setting. A common response is to reuse information from previously trained models through transfer, distillation, ensemble methods, or federated training instead of learning each target task from random initialization. The literature on these mechanisms is fragmented, and published comparisons are hard to interpret because tasks, baselines, and compute budgets differ. This paper presents a PRISMA-guided systematic review of empirical studies on pretrained knowledge reuse in DRL. Starting from 589 records retrieved from IEEE Xplore, the ACM Digital Library, and citation tracing, we screened 570 unique records and assessed 89 full texts. After applying the final eligibility criteria, 15 empirical studies remained in the main synthesis. We analyzed them qualitatively across three factors: source-target similarity, diversity among reused models, and the fairness of comparisons against from-scratch baselines. Three patterns recur across the surviving corpus. First, positive results are concentrated in settings where source and target tasks share substantial structure or where the method includes an explicit gating or alignment mechanism. Second, evidence for ensembles and federated aggregation is promising but sparse and mostly limited to narrow settings. Third, compute-matched comparisons are rare, which weakens claims about efficiency gains over stronger single-agent baselines. The paper contributes a narrower and internally consistent review scope, a study-level synthesis of empirical evidence, and a provisional independence spectrum that should be treated as a hypothesis for future benchmarking rather than a validated metric.

Ujjwal Patil, Javad Ghofrani

Department of Computer Science Hochschule Bonn-Rhein-Sieg (H-BRS) Sankt Augustin, Germany

Reinforcement Learning Knowledge Distillation

May 3, 2026

Bridging the Gap Between Average and Discounted TD Learning

The analysis of Temporal Difference (TD) learning in the average-reward setting faces notable theoretical difficulties because the Bellman operator is not contractive with respect to any norm. This complicates standard analyses of stochastic updates that are effective in discounted settings. Although a considerable body of literature addresses these challenges, existing theoretical approaches come with limitations. We introduce a novel algorithm designed explicitly for policy evaluation in the average-reward setting, utilizing sampling from two Markovian trajectories. Our proposed method overcomes previous limitations by guaranteeing convergence to the unique solution of a properly defined projected Bellman equation. Notably, and in contrast to earlier work, our convergence analysis is uniformly applicable to both linear function approximation and tabular settings and does not involve explicit dimension-dependent terms in its convergence bounds. These results align with what is known to hold in the discounted setting. Furthermore, our algorithm achieves improved dependence on the problem's condition number, reducing the sample complexity from quartic, as in prior literature, to quadratic scaling, and thus matching the efficiency seen in the discounted setting.

Haoxing Tian, Zaiwei Chen, Ioannis Ch. Paschalidis +1

Department of Electrical and Computer Engineering Boston University · Edwardson School of Industrial Engineering Purdue University

Reinforcement Learning Average-Reward RL Stochastic Approximation Policy Evaluation Temporal-Difference Learning Linear Function Approximation

May 3, 2026

Enhanced LLM Reasoning by Optimizing Reward Functions with Search-Driven Reinforcement Learning

Mathematical reasoning is a key benchmark for large language models. Reinforcement learning is a standard post-training mechanism for improving the reasoning capabilities of large language models, yet performance remains sensitive to the design of the reward function that drives policy optimization. This paper introduces a search-driven framework that treats the reward specification itself as an object of optimization. The setting of interest is one in which the base model is held fixed and the reward specification is the primary remaining design lever. Candidate reward functions are generated by a frontier language model, validated automatically, screened through 500-step Group Relative Policy Optimization (GRPO) training runs on a Llama-3.2-3B-Instruct base model with Low-Rank Adaptation (LoRA), and ranked by F1 on the GSM8K test set. Ranked summaries from prior rounds are then fed back into the next round of generation. Over five rounds, the search produces 50 candidate rewards. The mean F1 rises from 0.596 in Round 1 to 0.632 in Round 5, and the top individual reward reaches F1 = 0.787. Seven ensemble configurations of top-ranked rewards are evaluated. The best ensemble achieves F1 = 0.795 (95% bootstrap CI [0.756, 0.832]) and accuracy 0.660 [0.635, 0.686], a 0.19 absolute F1 gain over a base-rewards-only GRPO baseline (F1 = 0.609). Pairwise McNemar tests with Bonferroni correction show all five-or-more-reward configurations are statistically indistinguishable at α = 0.05/21. A three-seed re-training of the best ensemble yields F1 of 0.785. A randomly drawn 5-reward control collapses to F1 = 0.047, which shows that the ranked-feedback loop, not the additive signal of having more rewards, drives the gain.

Arash Ahmadi, Sarah Sharif, Yaser +1

Mike · School of Electrical and Computer Engineering, University of Oklahoma, Norman, OK 73019 USA · Intelligent Neuromorphic and Quantum Understanding for Innovative Research and Engineering (INQUIRE) Laboratory, University of Oklahoma, Norman, OK, USA

Reinforcement Learning Large Language Model-Guided Optimization RL for Language Model Reasoning Reward Design for RL

May 3, 2026

AdamO: A Collapse-Suppressed Optimizer for Offline RL

Offline reinforcement learning (RL) can fail spectacularly when bootstrapped temporal-difference (TD) updates amplify their own errors, driving the critic toward extreme and unusable Q-values. A key counterintuitive insight of this work is that collapse is not only a property of the backup rule or network architecture: optimizer dynamics themselves can directly trigger or suppress instability. From a control-theoretic viewpoint, we model offline TD learning as a feedback system and analyze Adam-based critic updates. This yields a necessary and sufficient condition for stability of the induced local update dynamics: within the regime we analyze, these dynamics are stable if and only if the spectral radius of the corresponding update operator is strictly below one. Further analysis suggests that standard Adam updates can inadvertently distort the parameter geometry, motivating explicit orthogonality constraints to prevent TD error amplification. To this end, we propose AdamO, an Adam-based optimizer with a decoupled orthogonality correction regulated by a strict task-alignment budget. We prove that this design theoretically guarantees worst-case task safety and preserves Adam's continuous-time dissipative dynamics. Empirically, AdamO is broadly compatible with diverse offline RL baselines, improving stability and returns across a broad suite of benchmarks.

Nan Qiao, Sheng Yue, Shuning Wang +1

1Tsinghua University, China · 2Central South University, China · 3Sun Yat-sen University, China

Reinforcement Learning Orthogonal Optimization RL Stability Temporal-Difference Learning Offline RL Neural Network Stability

May 3, 2026

MOIRA: Policy Memory and Belief Revision for Language-Model Agents in Non-Stationary Pair Trading

Many sequential decision-making problems exhibit hierarchical structure, where high-level semantic choices constrain downstream actions and feedback is delayed and ambiguous. Learning in such settings is challenging due to credit assignment: performance degradation may arise from flawed abstractions, suboptimal execution, or their interaction. We study this challenge through pair trading, a domain that naturally combines long-horizon semantic reasoning for asset pair selection with short-horizon execution under partial observability. We formulate pair trading as a hierarchical reinforcement learning problem and propose a language-driven optimization framework in which both high-level and low-level policies are parameterized by large language models (LLMs) and optimized exclusively through prompt updates. Our approach leverages pretrained LLMs as hierarchical policies and uses trajectory- and episode-level textual feedback to adapt abstractions and execution without gradient-based fine-tuning. By explicitly separating abstraction selection from execution, the framework reduces non-stationarity across hierarchical levels and enables targeted adaptation under delayed feedback. Experiments on real-world market data show consistent improvements over traditional and LLM-based baselines, demonstrating the effectiveness of language-driven hierarchical reinforcement learning.

Polydoros Giannouris, Yuechen Jiang, Lingfei Qian +5

The University of Manchester Manchester, UK · The Fin AI USA · Harvard University Cambridge, MA, USA +1

Reinforcement Learning Hierarchical RL Prompt Optimization Algorithmic Trading

May 3, 2026

Quality-Aware Exploration Budget Allocation for Cooperative Multi-Agent Reinforcement Learning

Cooperative multi-agent reinforcement learning (MARL) requires agents to discover joint strategies in a combinatorially large state-action space, yet effective coordination configurations are exceedingly rare. Intrinsic motivation, which augments task rewards with novelty bonuses, is a popular approach for driving exploration, but its effectiveness hinges on the exploration intensity

β

, where too large a value overwhelms the task signal and causes coordination collapse, while too small a value prevents discovery of rare strategies. We address two complementary challenges: adapting

β

globally over training, and allocating the exploration budget across agents whose intrinsic reward signals vary in reliability. Our framework combines a return-conditioned sigmoid schedule (RCB) for global intensity control with a per-agent Reward Signal Quality (RSQ) metric that concentrates the exploration budget on agents with reliable signals. The core insight is that agents receiving noisy intrinsic rewards should explore less aggressively, and this allocation can be determined automatically from signal-to-noise statistics. Successor Distance (SD), a quasimetric intrinsic reward, naturally produces distinguishable per-agent signal quality, completing the framework with convergence and ordering preservation guarantees. On seven cooperative benchmarks (MPE, SMAX, MABrax), our method achieves top-tier returns across all environments.

Dahyun Oh, Minhyuk Yoon, H. Jin Kim

Department of Aerospace Engineering, Seoul National University, Seoul, Republic of Korea

Multi-Agent Coordination Intrinsic Reward Methods Reinforcement Learning Intrinsic Motivation RL Exploration Cooperative MARL

May 3, 2026

Remote Action Generation: Remote Control with Minimal Communication

We address the challenge of remote control where one or more actors, lacking direct reward access, are steered by a controller over a communication-constrained channel. The controller learns an optimal policy from observed rewards and communicates action guidance to the actors, which becomes demanding for large or continuous action spaces. To achieve rate-efficient communication throughout this interactive learning and control process, we introduce a novel framework leveraging remote generation. Instead of transmitting full action specifications, the controller sends minimal information, enabling the actors to locally generate actions by sampling from the controller's evolving target policy. This guided sampling is facilitated by an importance sampling approach. Concurrently, the actors use the received guidance as supervised learning data to learn the controller's policy. This actor-side learning improves their local sampling capabilities, progressively reducing future communication needs. Our solution, Guided Remote Action Sampling Policy (GRASP), demonstrates significant communication reduction, achieving an average 12-fold data reduction across all experiments (50-fold for continuous action spaces) compared to direct action transmission, and a 41-fold reduction compared to reward transmission.

Szymon Kobus, Deniz Gündüz

Electrical and Electronic Engineering Imperial College London

RL Control Reinforcement Learning Feedback Control

May 3, 2026

Selector-Guided Autonomous Curriculum for One-Shot Reinforcement Learning from Verifiable Rewards

Recently, Reinforcement Learning from Verifiable Rewards (RLVR) has been established as a highly effective technique for augmenting the math reasoning skills of Large Language Models (LLMs) based on a single instance. Current state-of-the-art 1-shot RLVR models adopt heuristics for selecting instances, mostly based on historical variance in rewards, which we find to be inherently misleading as a measure of transferability value. In this paper, we propose a Selector-Guided Autonomous Curriculum (SGAC) approach, which employs a learnable selector model on a multi-dimensional feature space consisting of success probability, reward variance, output disagreement (entropy), and semantic difficulty level, instead of the static reward variance heuristic. In our empirical evaluation on pools of candidate problems, we observed that output disagreement, rather than reward variance, is the strongest predictor of reasoning gains in subsequent iterations. Leveraging this finding, we develop an autonomous curriculum algorithm for dynamically siphoning candidate problems from a large pool, ranking them by the learned selector, and running micro-bursts of 1-shot GRPO. Our framework is evaluated using the Hendrycks MATH benchmark, with the Qwen2.5-Math-1.5B model serving as the baseline. Our framework obtains an accuracy of 68.0% on the hold-out dataset, which is better than the accuracy obtained from the state-of-the-art model, 64.0%, as well as the 1-shot RLVR checkpoint proposed by Wang et al., which achieved an accuracy of 66.0%. The results confirm that entropy-based intelligent data curation leads to strict reasoning improvement over static training methods, particularly in severely limited data conditions.

Rudray Dave, Vedang Dubey, Smit Deoghare +1

Department of Artificial Intelligence Sardar Vallabhbhai National Institute of Technology Surat, India

Reinforcement Learning RL for Language Model Reasoning Curriculum RL Reinforcement Learning with Verifiable Rewards

May 3, 2026

MAGIC: Multi-Step Advantage-Gated Causal Influence for Multi-agent Reinforcement Learning

A key challenge in multi-agent reinforcement learning (MARL) lies in designing learning signals that effectively promote coordination among agents. Designing such signals requires estimating how one agent's current action affects its teammates over future interaction steps. To address this, we introduce Multi-step Advantage-Gated Interventional Causal MARL (MAGIC), a framework that estimates multi-step action effects between agents and selectively converts them into intrinsic rewards. MAGIC uses counterfactual action interventions to compare teammate futures under factual and counterfactual branches, and introduces a gate based on advantage to direct exploration toward beneficial behaviors aligned with the task goal. Experiments on Multi-Agent Particle Environments (MPE) and StarCraft micromanagement benchmarks (SMAC and SMACv2) show that MAGIC consistently outperforms leading prior methods, with average relative final performance improvements of 26.9% and 10.1%, respectively.

Haohan Yu, Jinmiao Cong, Shengzhi Wang +2

1Dalian University of Technology · 2Microsoft, Beijing, China

Multi-Agent Coordination Intrinsic Reward Methods Reinforcement Learning Cooperative MARL Causal RL Reward Design for RL+1

May 3, 2026

Adversarial Imitation Learning with General Function Approximation: Theoretical Analysis and Practical Algorithms

Adversarial imitation learning (AIL), a prominent approach in imitation learning, has achieved significant practical success powered by neural network approximation. However, existing theoretical analyses of AIL are primarily confined to simplified settings, such as tabular and linear function approximation, and involve complex algorithmic designs that impede practical implementation. This creates a substantial gap between theory and practice. This paper bridges this gap by exploring the theoretical underpinnings of online AIL with general function approximation. We introduce a novel framework called optimization-based AIL (OPT-AIL), which performs online optimization for reward learning coupled with optimism-regularized optimization for policy learning. Within this framework, we develop two concrete methods: model-free OPT-AIL and model-based OPT-AIL. Our theoretical analysis demonstrates that both variants achieve polynomial expert sample complexity and interaction complexity for learning near-expert policies. To the best of our knowledge, they represent the first provably efficient AIL methods under general function approximation. From a practical standpoint, OPT-AIL requires only the approximate optimization of two objectives, thereby facilitating practical implementation. Empirical studies demonstrate that OPT-AIL outperforms previous state-of-the-art deep AIL methods across several challenging tasks.

Tian Xu, Zhilong Zhang, Zexuan Chen +3

National Key Laboratory for Novel Software Technology and School of Artificial Intelligence, Nanjing University · School of Mathematics, Nanjing University

Reinforcement Learning Adversarial Imitation Learning Imitation Learning

May 3, 2026

TRIMMER: A New Paradigm for Video Summarization through Self-Supervised Reinforcement Learning

The rapid growth of video content across domains such as surveillance, education, and social media has made efficient content understanding increasingly critical. Video summarization addresses this challenge by generating concise yet semantically meaningful representations, but existing approaches often rely on expensive manual annotations, struggle to generalize across domains, and incur significant computational costs due to complex architectures. Moreover, unsupervised and weakly supervised methods typically underperform compared to supervised counterparts in capturing long-range temporal dependencies and semantic structure. In this work, we propose TRIMMER (Temporal Relative Information Maximization for Multi-objective Efficient Reinforcement), a novel self-supervised reinforcement learning framework for video summarization. TRIMMER operates in two stages: it first learns robust representations via self-supervised learning and then performs spatio-temporal decision making through reinforcement learning guided by information-theoretic reward functions. Unlike prior approaches that rely on similarity-based objectives, our method introduces entropy-based metrics to capture higher-order temporal dynamics and semantic diversity, while computing rewards directly over selected frame indices to improve computational efficiency. Extensive experiments on standard benchmarks demonstrate that TRIMMER achieves state-of-the-art performance among unsupervised and self-supervised methods, while remaining competitive with leading supervised approaches, highlighting its effectiveness for scalable and generalizable video summarization.

Pritam Mishra, Coloma Ballester, Dimosthenis Karatzas

Universitat Pompeu Fabra, Barcelona, Spain · Universitat Autònoma de Barcelona, Barcelona, Spain

Reinforcement Learning Video Summarization Self-Supervised RL Video Representation Learning

May 2, 2026

MIRL: Mutual Information-Guided Reinforcement Learning for Vision-Language Models

Vision-Language Models (VLMs) frequently suffer from visual perception errors and hallucinations that compromise answer accuracy in complex reasoning tasks. Reinforcement Learning with Verifiable Rewards (RLVR) offers a promising solution by optimizing policies using answer correctness signals. Despite their effectiveness, prevailing RLVR methods face two critical limitations. First, much of the sampling budget is wasted on trajectories doomed to fail due to early visual description errors. Second, sparse rewards cannot distinguish whether failures stem from visual perception or reasoning stages. We introduce MIRL, a decoupled framework that addresses both limitations by leveraging mutual information (MI) between generated descriptions and visual inputs as a cheap pre-screening signal. This enables intelligent budget allocation toward high-potential trajectories via forking, while decoupled training provides independent MI-based rewards for visual perception optimization, resolving reward blindness. Experiments on six vision-language reasoning benchmarks demonstrate that MIRL achieves 70.22% average accuracy and successfully surpasses the performance of sampling 16 complete trajectories using only 10 pre-samples with top-6 selection (25% fewer complete trajectories). Our code is available at: https://anonymous.4open.science/r/mirl-main/.

Yin Zhang, Jiaxuan Zhao, Zonghan Wu +5

School of Mathematics, Tianjin University, Tianjin, China · Institute of Information Engineering, Chinese Academy of Sciences, Beijing, China · Shanghai Advanced Institute of Finance (SAIFS), East China Normal University, Shanghai, China +4

VLM Reasoning Vision-Language Models Reinforcement Learning RL for VLMs Reinforcement Learning with Verifiable Rewards Visually Grounded Reasoning+1

May 2, 2026

PACE: Parameter Change for Unsupervised Environment Design

Unsupervised Environment Design (UED) offers a promising paradigm for improving reinforcement learning generalization by adaptively shaping training environments, but it requires reliable environment evaluation to remain effective. However, existing UED methods evaluate environments using indirect proxy signals such as regret, value-based errors, or Monte Carlo, which suffer from bias, high variance, or substantial computational overhead and fail to reflect agent realized learning progress. To address these limitations, we propose Parameter Change Environment Design (PACE), which evaluates an environment through the policy parameter change induced by training on that environment, directly grounding environment selection in realized learning progress. Specifically, PACE assigns environment value using a first-order approximation of the policy optimization objective, where the improvement induced by an environment is proportional to the squared L2 norm of the corresponding parameter update, enabling low-variance and computation-efficient evaluation without additional rollouts. Experiments on MiniGrid and Craftax show that PACE consistently outperforms established UED baselines, achieving higher IQM and smaller Optimality Gap on OOD evaluations, including an IQM of 96.4% and an Optimality Gap of 17.2% on MiniGrid.

Fang Yuan, Quanjun Yin, Siqi Shen +5

College of Systems Engineering, National University of Defense Technology, Changsha, China · Test Center, National University of Defense Technology, Xi’an, China · State Key Laboratory of Digital Intelligent Modeling and Simulation, Changsha, China +2

Reinforcement Learning