Reinforcement Learning

Vision-Language Models (VLMs) often achieve high performance on benchmarks while remaining "black boxes", yet they remain prone to hallucination or rely on superficial shortcuts. In this work, we propose a framework designed to enhance both performance and interpretability through De-compositional Evidence Grounding. Unlike monolithic inference approaches, our approach forces the model to decompose a global query into a sequence of atomic sub-questions, each requiring an explicit sub-answer and critically a localized evidence bounding box. By grounding intermediate logical steps (e.g. identifying a container, analyzing liquid properties, and assessing environmental context) in specific visual regions, we construct a structured reasoning path that mirrors human-like deduction. This allows the final answer to emerge as a logical consequence of verified visual facts rather than a statistical guess.

Reinforcement learning (RL) has become indispensable for pushing Vision-Language-Action Models (VLAs) beyond static imitation learning. However, existing RL methods typically require external environmental feedback, relying on predefined success signals to guide policy updates. In this work, we show that VLA models possess useful internal evaluative capabilities: in discrete-action VLAs, trajectories with higher generation confidence are significantly more likely to succeed. Based on this observation, we introduce T^2VLA (Test-time VLA), an architecture-agnostic test-time RL framework that enables VLA models to achieve self-bootstrapping policy improvement. Instead of relying on external rewards, T^2VLA leverages trajectory-level similarity to high-confidence expert demonstrations as an intrinsic reward signal. In addition, we propose a Confidence-Driven Dual Expert Bootstrapping mechanism, which dynamically balances a Local Pseudo-Expert for exploration and a Global Expert Pool for training stability. Extensive experiments on the LIBERO and RoboTwin benchmarks show that T^2VLA consistently outperforms supervised baselines and approaches oracle RL performance with ground-truth rewards, achieving effective improvement without external reward feedback. Furthermore, T^2VLA adapts to distinct VLA paradigms, including both OpenVLA-OFT and the pi series.

Deep Reinforcement Learning (DRL) has achieved significant success in robotics and autonomous systems, yet remains vulnerable to adversarial perturbations that can severely degrade performance. Research in adversarial reinforcement learning is often limited by fragmented implementations, inconsistent evaluation protocols, and poor reproducibility. To address these challenges, we present \textbf{RoAd-RL}, an open-source benchmarking framework that provides unified abstractions for policies, attacks, defenses, and robustness metrics, together with reproducible evaluation pipelines and seamless integration with Stable-Baselines3 and Gymnasium. We evaluate DQN, PPO, and SAC agents in LunarLander and Highway-v0 under 192 attack-defense configurations. Results reveal substantial variations in robustness across environments and show that some commonly used defenses can be more detrimental than the attacks they aim to mitigate, while temporal smoothing consistently achieves strong performance. RoAd-RL establishes a standardized benchmark for adversarial reinforcement learning research and is publicly available at https://pypi.org/project/road-rl.

In complex continuous-control reinforcement learning tasks, multimodal optimal actions often coincide with uncertain, multimodal return distributions, making reliable value estimation and multimodal exploration challenging. Existing value estimation methods using unimodal Gaussians restrict expressiveness and yield biased estimates. Recent generative policies can represent multimodal actions but often collapse to a few modes and under-explore high-value areas of the action space. Motivated by these challenges, we propose Dual-Flow RL, a unified actor-critic framework that jointly models a continuous return distribution and a multimodal policy distribution using conditional flow matching (CFM). This design supports reliable value estimation and sustained multimodal exploration. To further enhance exploration, we introduce an Entropy-Covariance Exploration Regulator (ECER) that enables state-aware exploration regulation leveraging policy entropy and action-uncertainty covariance. Experiments on DeepMind Control Suite and Humanoid-Bench show that Dual-Flow RL achieves state-of-the-art performance on most tasks, significantly outperforming prior diffusion-based and flow-based methods.

Reinforcement Learning from Human Feedback (RLHF) for Large Language Models increasingly relies on critic-free methods as a practical alternative to actor--critic training. Despite their simplicity, existing critic-free approaches propagate a trajectory-level learning signal uniformly across all tokens in a trajectory. This requires full-trajectory policy updates for every rollout, leading to substantial optimization cost for long reasoning traces, even though intermediate prefixes often contain enough information to largely determine the final outcome. We propose Prefix-Sampling Proximal Policy Optimization (PS-PPO), a compute-efficient critic-free method for RLHF that exploits this temporal redundancy. PS-PPO introduces a prompt-conditioned cutoff distribution and samples a cutoff timestep for each trajectory. During the update pass, PS-PPO backpropagates only through the sampled prefix of each trajectory and applies an importance-weighting correction so that the resulting truncated gradient estimator remains unbiased with respect to the full-trajectory objective. Experiments on mathematical reasoning and RLHF benchmarks show that PS-PPO achieves large reductions in training compute and peak GPU memory, while maintaining accuracy comparable to strong critic-free baselines.

Reinforcement learning (RL) has gained growing attention in large language model (LLM) post-training, yet RL training remains fragile and can suffer from instability or collapse. One vital cause is training-inference mismatch: LLM adopts separate inference and training engines for generation efficiency and training precision, which in practice exhibits inconsistent probabilities for the same trajectories on training and inference sides, even with synchronized model parameters. This naturally induces a special type of off-policyness ever existing and poisoning the training. Prior works have made various efforts in addressing the off-policyness to stabilize the training policies under the mismatch. In this paper, we point out the objective misalignment neglected by existing works that an effective update to the policy in the training engine not necessarily ensures the improvement of the inference policy, i.e., the one used in deployment. To this end, we propose a new policy optimization objective for LLM RL, named Monotonic Inference Policy Improvement (MIPI). Following this principle, we introduce Monotonic Inference Policy Update (MIPU), a two-step LLM RL framework that constructs sampler-referenced candidate updates and selectively accepts synchronized candidates using an inference-side gap proxy. Experiments conducted on two model scales under high mismatch show that MIPU improves average reasoning performance and training stability.

World 1-1 of Super Mario Bros is widely celebrated as a masterclass in game design: its progressive structure is credited with teaching players core mechanics through the level itself. We ask whether that structure is empirically measurable using reinforcement learning. We implement World 1-1 from scratch as a fully discrete environment and compare four algorithms -- Q-Learning, SARSA, Monte Carlo, and Deep Q-Network (DQN) -- across three progressively complex versions of the same level. Monte Carlo emerges as the strongest agent (94.9% $\pm$ 1.5% win rate), outperforming DQN (76.4% $\pm$ 3.4%) by learning to maximize intermediate rewards along winning paths rather than taking the most direct route. We then use Monte Carlo in a curriculum experiment permuting World 1-1's six canonical segments across twelve conditions. Canonical ordering converges fastest, achieves the highest learning efficiency, and is the only condition with zero catastrophic failures; no random permutation matches all three criteria simultaneously. These results provide, to the best of our knowledge, the first empirical validation that World 1-1's canonical design encodes genuine pedagogical structure: one that measurably accelerates learning and cannot be replicated by chance.

Group Relative Policy Optimization (GRPO) is a default recipe for process-supervised reinforcement learning of LLM reasoners, and dense process supervision -- via learned process reward models (PRMs) or on-policy-distillation KL signals -- is a common way to densify its otherwise weak outcome reward. Layering such a step-level signal on top of GRPO's group-standardized advantage, however, exposes three structural pathologies: \emph{channel contamination} between the pooled process, outcome, and format streams at group standardization; \emph{resolution mismatch} between the granularity of the process signal and the granularity of the logical decisions being credited; and a \emph{cumulative trap} by which GRPO's return-to-go sum surfaces either length inflation or truncated exploration depending on the sign regime of the signal. We propose \textbf{PASS} (\emph{Process Advantage Signal Shaping}), a compact middleware that sits between any scalar step-level process signal and GRPO's clipped surrogate and addresses the three pathologies in turn: \emph{Advantage Fusion} standardizes the three streams independently within each group, \emph{Chunk-by-Value} derives value-homogeneous chunks from the signal itself and broadcasts credit within each chunk, and \emph{Divide-Length} converts the cumulative objective into an average-value-density score. We validate PASS across two domains and two process-signal paradigms -- a learned PRM on mathematical reasoning and an on-policy-distillation KL signal (with a generalized variant) on multi-hop question answering -- and under two group-standardization operators. In every regime PASS delivers a consistent pass@1 gain over the corresponding GRPO baseline.

Autonomous navigation in unknown environments requires a robot to efficiently reach a predefined goal while exploring without prior maps. Although progress has been made in this area, most existing works still rely on traditional planning methods with hand-crafted rules, while learning-based methods often suffer from limited environmental memory and challenges in simulation-to-real (sim-to-real) transfer. To overcome these limitations, we propose a Contextual-memory Oriented Reinforcement-learning (CORE) planner for robot navigation in unknown environments. The proposed CORE planner effectively combines the core advantages of traditional and learning-based methods. Specifically, our method uses a sparse visibility graph for structured environment representation, reducing the computational overhead of dense grid maps, and employs a Transformer network to achieve a holistic environmental understanding, thereby significantly improving navigation efficiency. Moreover, we introduce a visibility graph-based graph sparsification method and a contextual memory mechanism, which alleviates local optima and enhances computational performance in large-scale scenes. Finally, our approach achieves zero-shot sim-to-real transfer after training solely on image-based environments, requiring no fine-tuning. Experimental results show that CORE Planner consistently outperforms state-of-the-art methods, including the traditional FAR Planner and all learning-based baselines, across representative environments, reducing travel distance by 13% over traditional FAR Planner and by up to 48% relative to learning-based baselines, with larger gains observed in more complex environments. In real-world scenarios, CORE successfully navigates without human intervention, showcasing zero-shot sim-to-real transfer. Code is available at https://github.com/BBD00/core_planner.

Reinforcement learning agents can exploit misspecified reward signals to achieve high apparent returns while failing on the intended objective, a failure mode known as reward hacking. Existing practical defenses typically constrain policy updates to stay near a known safe reference, creating a tension between suppressing hacking and permitting legitimate improvement. We propose Modification-Considering Value Learning (MCVL), which operationalizes the theoretical idea of current utility optimization for standard value-based RL. MCVL wraps an off-policy learner and treats each incoming transition as a candidate modification: it forecasts two training paths, one that includes the transition and one that does not, and scores both with a frozen bootstrapped-return estimator derived from a learned reward model and value function. The transition is admitted only if inclusion does not decrease the score. We formalize conditions under which this filtering is both safe and permissive, and instantiate MCVL with DDQN and TD3. Across four safety-relevant gridworlds and three modified MuJoCo continuous-control tasks with diverse hacking mechanisms, MCVL mitigates reward hacking while continuing to improve the intended objective. Project website: ktolnos.github.io/mcvl/.

Critic-free reinforcement learning with verifiable rewards (RLVR), exemplified by Group Relative Policy Optimization (GRPO), avoids training a value function (critic) and reduces memory and compute overhead relative to critic-based PPO pipelines for aligning large language models. However, GRPO-style advantage estimation depends on prompt-local (within-prompt-group) reward statistics and can be unstable. In particular, when all rollouts in a prompt group receive identical rewards, the within-group reward variance becomes zero, and group normalization yields zero advantages for that group, impeding learning in cold-start regimes with binary verifiers. We introduce BV-Blend, a critic-free framework that stabilizes advantage estimation by combining prompt-local on-policy statistics with semantic-cluster-conditioned historical moments. BV-Blend maintains EMA-tracked reward moments for each cluster, derives a confidence weight from a standard error of the mean (SEM) proxy, and uses this weight to blend historical and prompt-local baseline and variance statistics into a standardized advantage for PPO-style clipped updates. Experiments on verifiable reasoning benchmarks show that BV-Blend improves training stability and performance, and remains robust in regimes where group-normalized methods may stall.

Neuromorphic and edge computing research has focused on reducing the inference cost of neural network controllers, yet in physical closed-loop systems the actuator can rival or exceed an efficient controller in energy. An efficient controller is therefore necessary but not sufficient, because the actuator becomes the cost worth reducing once inference no longer dominates it. Here, we introduce energy-aware learning, an approach that incorporates actuator energy directly into the reinforcement learning reward, and demonstrate it in closed-loop deep brain stimulation (DBS) for Parkinson's disease. A deep spiking Q-network, trained in a biophysical cortico-basal ganglia-thalamic circuit model, learns to suppress pathological alpha-beta oscillations by 45.2% while reducing stimulation charge by 80.0% relative to continuous DBS. Sparsity-constrained knowledge distillation compresses the policy onto the SynSense XyloAudio 3 neuromorphic processor at 0.52 mW inference power, yielding 28.1x lower energy per inference than an equivalent artificial neural network on conventional edge hardware. By co-optimizing stimulation energy and inference efficiency, the framework addresses both major power demands in implantable neuromodulation.

Reinforcement learning with verifiable rewards (RLVR) has significantly improved the reasoning capability of large language models, reaching expert or even superhuman performance in domains such as competition math. However, whether weaker agents and humans can actually harness this capability is far less certain, with RLVR documented to drift reasoning toward idiosyncratic patterns such as poor readability and language mixing. Tandem training is a recently introduced paradigm that targets this compatibility problem: a trained, stronger senior co-generates each rollout with a frozen, weaker junior, and the two are rewarded as a team, so the senior is pushed to reason in ways the junior can follow. Yet this paradigm has so far been demonstrated only in proof-of-concept settings, leaving open whether it scales to the long chains of thought of the modern RLVR pipeline. In this work, we propose Tandem Reinforcement Learning (TRL), which carries the tandem training paradigm into RLVR. In TRL, the senior and a frozen junior alternate stochastically to co-generate the reasoning, the resulting generation is rewarded, and the standard GRPO loss is applied to the senior. Training Qwen3-4B-Instruct on competition math, we find that TRL matches vanilla GRPO on solo reasoning capability while three properties emerge together from the same rollout structure: stronger handoff robustness with the junior, reduced distributional drift from the junior, and a chain-of-thought more legible to the junior. Our results demonstrate a promising route for RLVR with practical payoffs in multi-model communication and human compatibility.

We propose KL-Coupled Policy Regularization (KCPR), a policy coordination framework for Reward-Punishment Reinforcement Learning (RPRL). Based on KCPR, we derive KL-Coupled Soft Optimality (KCSO) and develop its deep realization, klDMP. Unlike existing RPRL approaches that optimize reward-seeking and punishment-related policies largely independently, KCPR enables direct interactions between companion policies by treating each as a dynamically learned prior for the other. KCSO yields coupled soft-optimal policies and KL-regularized Bellman operators, allowing reward and punishment information to jointly influence value propagation. To improve learning stability, we introduce a companion-prior softening mechanism and evaluate separate replay-buffer designs for balancing reward- and punishment-related experience. Experiments in grid-world and Gazebo robotic navigation tasks demonstrate that klDMP improves safety and learning stability while maintaining competitive task performance compared with DQN, SQL and softDMP. These results suggest that policy-level coordination provides an effective mechanism for integrating multiple behavioral objectives and may serve as a useful design principle for reinforcement learning systems with interacting motivational processes.

One goal in reinforcement learning (RL) research is to understand general-purpose sequential decision-making, using benchmark simulators as a proxy for learning in deployment settings. When running experiments, however, the goal of achieving high performance in the simulator can mutate into focusing exclusively on solving the simulator. To achieve high scores, researchers may adopt solutions exclusively meant for solving simulators, rather than learning while the agent is deployed outside a simulator. Solving simulators is also worthy of investigation, but it is a fundamentally different RL research question. In this paper, we argue that RL researchers need to distinguish between two use cases of simulators: solving simulators and using simulators as a proxy for learning in deployment. We first discuss how these two use-cases are importantly different, in terms of constraints on how the agent can use the simulator, which algorithms are appropriate, and which evaluation metrics are appropriate. We then highlight several issues and misleading conclusions that can occur by not making the distinction between these two settings clear, supported with examples and simple experiments. This work is a call to the community to begin clearly distinguishing how they are using simulators in their work, hopefully sparking further discussion on which empirical practices work best in each setting.

Reinforcement learning from human feedback (RLHF) in production does not always have a synchronous reward signal. Code-execution verifiers, slow judge ensembles, and queued human review can return several gradient steps after the rollout that produced them, breaking the synchronous-reward assumption underlying standard PPO. We address this gap with Retroactive Advantage Correction (RAC): each pending slow completion is queued, aged through a non-negative kernel, and reinjected as a clipped residual into the next optimiser step's advantage. We prove that under an unbiased clipped importance ratio, the cumulative RAC correction is exactly unbiased when the effective delay kernel reinjects all of its mass, and carries a bias linear in the unreinjected fraction otherwise; at the no-delay identity kernel it reduces to V-trace. On a tabular Markov decision process (MDP) proof-of-concept, RAC reduces the closed-form policy bias by up to 47.9x at the two-slow-channel configuration, beating wait-for-slow at lower wall-clock cost. RAC integrates with PPO and GRPO through a two-line reward-manager patch.

Reinforcement learning with verifiable rewards (RLVR) for training LLMs typically rely on ground-truth answers to assign rewards, limiting their applicability to tasks where the ground-truth solution is unknown. We introduce a \textbf{R}anking-\textbf{i}nduced \textbf{VER}ifiable framework (RiVER) that trains LLMs on score-based optimization tasks without ground-truth solutions, using deterministic execution feedback as continuous-valued supervision. When applying group-relative RL to such continuous rewards, we identify two key challenges: \emph{scale dominance}, where uncalibrated score magnitudes across test instances distort policy updates, and \emph{frequency dominance}, where repeatedly sampled suboptimal solutions can outweigh rare but stronger candidates. RiVER addresses these challenges with calibrated reward shaping that uses instance-wise comparisons and emphasizes top-ranked solvers while retaining bounded feedback for other valid solutions. We train on 12 AtCoder Heuristic Contest tasks and evaluate on Algorithm Engineering Benchmark (ALE-Bench), LiveCodeBench, and USACO. RiVER advances Qwen3-8B and GLM-Z1-9B-0414 by 8.9% and 9.4% in ALE rating rank. More importantly, despite training exclusively on score-based tasks without any ground-truth solutions, RiVER also improves the backbones across exact-solution benchmarks such as LiveCodeBench and USACO by an absolute average improvement of 2.4% and 3.5%. By contrast, baselines trained with raw execution scores improve ALE rating but fail to transfer to exact-solution benchmarks. These results suggest that score-based optimization tasks, combined with proper reward calibration, can serve as effective training environments for general coding ability without ground-truth solutions.

Large language model (LLM) post-training for reasoning increasingly relies on reinforcement learning with verifiable rewards (RLVR), where models learn from ground-truth feedback on mathematical, logical, and scientific tasks. To enable flexible resource allocation and support heterogeneous training setups, modern RLVR systems adopt disaggregated architectures that decouple rollout generation and policy training across independent GPU pools. However, existing synchronous on-policy GRPO (Group Relative Policy Optimization) RLVR systems finish an entire rollout before starting training, leaving the trainer GPU pool idle while rollout is still ongoing. Asynchronous RL pipelines overlap the two stages, but at the cost of training on stale data. To address these challenges, we propose RolloutPipe, a post-training framework for disaggregated RLVR systems, which turns the fixed-weight rollout into a complete-group pipeline where trainable groups move to the trainer while later groups are still being generated. RolloutPipe achieves this through two techniques including complete-group pipelining (CGP) and frontier-group dispatch (FGD). CGP dispatches each trainable complete group to the trainer FIFO as soon as group materialization finishes, and FGD is an admission policy on the Rollout node that first admits requests for the frontier groups needed to form the next training batch, so that trainer-ready groups arrive earlier and more steadily. The design starts training before the rollout completes while maintaining on-policy correctness. Evaluated on Qwen3-1.7B across four reasoning and science benchmarks and twelve rollout settings, RolloutPipe shortens the rollout-to-train-end time by 30.7%-42.3%, and lowers the trainer waiting ratio by 37%-76% compared to Slime, a state-of-the-art rollout and training system.

Online reinforcement learning is widely used to align large language models (LLMs) with reward signals, yet training can be unstable under noisy or misspecified rewards. We identify a failure mode we call directional inconsistency: within a batch, a small set of high-reward rollouts induces representation-space preference directions that sharply disagree with the batch majority, resulting in high-variance and destabilizing updates. We propose geoalign, a lightweight plug-in for rollout curation in iterative policy optimization. Geoalign (i) forms within-prompt preference pairs, (ii) learns an online projector on per-rollout hidden states to concentrate reward-ordered displacement directions, and (iii) detects directionally inconsistent rollouts via their angular deviation from a batch consensus prototype and rectifies them with within-prompt stable alternatives. Geoalign is forward-pass only and adds negligible overhead. Across dialogue alignment with a learned reward model and mathematical reasoning with binary verified rewards, Geoalign improves final performance and reduces training oscillation, outperforming PF-PPO, PAR, PODS, and Seed-GRPO. These results suggest latent directional consensus as an effective reliability signal for online LLM RL.

Many real-world control problems involve hybrid discrete-continuous action spaces. For example, steering and signaling in autonomous driving, and aiming and firing in robotics or video-games. Despite real-world hybrid factorization and reinforcement learning framework support for complex action spaces (e.g., Gymnasium, PettingZoo, TorchRL, SeedRL, Mujoco, etc), the default environments within those frameworks often implement uniform action space configurations (LunarLander, Walker2D, Cheetah, SMAC, SUMO, Ant, Atari). Landmark hybrid-action benchmarks (RoboCup 2D HFO, SC2LE, Platform, CARLA, etc) are mostly heavyweight or archival implementations originating from papers which test one or a small number of competing factorization methods on one kind of control. This article provides a cross-sectional study of factorization methods [independent networks, shared encoder, VDN, QPLEX, Joint, Auto-Regressive] on each of three families of algorithms [PPO, SAC, DQN] across three action spaces [discretized, hybrid, continuous] over four lightweight environments [Platform, hybrid-LunarLander, Hybrid-Shoot, CoopPush]. Accounting for some invalid pairings such as joint-continuous, we are left with 220 configurations to analyze each method. We provide two new C++ parallel gymnasium and petting-zoo compliant environments [CoopPush, Hybrid-Shoot] to isolate particular challenges such as state-dependent inter-action dependence. Finally, we introduce VDN-PPO and PPO-MIX which use a branching critic to assign credit to multi-headed PPO. These variants out-perform all other tested PPO factorizations. Our results suggest that branching dueling architectures balance compute and performance most effectively, with Auto-Regressive actions reaching the highest performance overall and native continuous SAC outperforming discrete and hybrid algorithms, albiet both at increased computational cost.

Recently, zero-shot text-to-speech (TTS) has enabled high-fidelity and expressive speech synthesis, but it often fails to imitate unseen speaking styles from uncommon scenarios (e.g., crosstalk, dialects). Moreover, fine-tuning pretrained models requires large, high-quality datasets, limiting rapid personalization. We propose VoiceTTA, a reinforcement learning-based test-time adaptation (TTA) method that improves voice imitation of pretrained zero-shot TTS models. VoiceTTA introduces two style rewards based on coefficient-of-variation differences of F0 and energy, combined with speaker similarity and intelligibility (WER from a pretrained Whisper model), and optimizes learnable prefixes via group relative preference optimization (GRPO) in a flow matching-based model at inference time. Extensive experiments demonstrate substantial improvements on uncommon speech prompts, outperforming state-of-the-art baselines. Audio samples are available at https://voicetta.pages.dev/

Transfer learning improves policy learning efficiency by reusing knowledge from source tasks, providing a feasible paradigm for safe and efficient autonomous highway lane changing decision-making. Existing methods frequently encounter transfer mismatch induced by distribution shifts between source and target domains, leading to training oscillation and performance decline. Besides, target domain adaptation depends on exploratory interactions, which struggles to guarantee training safety in safety-critical lane changing cases. To tackle these limitations, this paper proposes a safe transfer reinforcement learning framework for autonomous highway lane changing. First, we design an adaptive teacher intervention mechanism based on instantaneous safety cost to restrain risky exploration and fade intervention strength progressively, with theoretical analysis on return bounds for mixed behavior policy. This intervention also produces dual-source samples for joint training. Second, a teacher-guided safe transfer module embeds action evaluation information of teacher policy into student learning via reward shaping to boost training safety and efficiency, with teacher guidance decaying as policy safety rises. Third, a teacher-guided weighted optimization mechanism adjusts sample weights in policy optimization using a likelihood ratio factor to stabilize transfer performance. Experiments under varied traffic densities and validations on real-world NGSIM dataset reveal that our method surpasses baseline approaches by over 52.2% in safety and 5.0% in learning efficiency. Results verify the efficacy and robustness of our safety-aware transfer strategy for autonomous highway lane changing under various traffic conditions.

This paper addresses model-free continuous-time mean-field control in a setting where the population dynamics evolve continuously according to an unknown McKean-Vlasov stochastic differential equation, while only discrete-time transition data are available. In the model-based formulation, policy evaluation is naturally described by a stationary Hamilton-Jacobi-Bellman equation on $\mathcal P_2(\mathbb R^d)$, but this equation involves the drift and diffusion coefficients of the controlled McKean-Vlasov dynamics, which are not identifiable when only discrete-time data are available. On the other hand, a direct reduction to a time-discrete Bellman equation avoids the non-identifiability issue but loses the differential equation structure. To bridge these two viewpoints, we introduce a Mean-Field-PhiBE (MF-PhiBE), which incorporates discrete-time transition information into a continuous-time PDE on the Wasserstein space. The MF-PhiBE replaces the unknown infinitesimal drift and covariance in the policy-evaluation equation by one-step estimators computed from data, while preserving the generator structure of the McKean-Vlasov HJB equation. We also derive a policy-gradient theorem for entropy-regularized randomized feedback policies, expressing the actor direction through an action-wise infinitesimal advantage and the score of the policy. Combining these two ingredients yields a model-free actor-critic method. We prove a first-order consistency estimate showing that the value induced by an optimal MF-PhiBE policy approximates the optimal continuous-time value with an error of order $Δt$. In the linear-quadratic case, we show our approximation achieves second-order accuracy with only one-step data. Numerical experiments on an LQR benchmark and a crowd-aversion problem illustrate the proposed framework.

Deliberating takes time. In real-time settings, that time is not free. Standard reinforcement learning (RL) sidesteps this as the environment waits indefinitely for the agent's decision. Instead, we study real-time RL environments where the environment progresses while waiting for the agent's action. Building on prior real-time formalizations, we introduce variable-delay real-time RL, where the agent chooses how long to deliberate at each decision point since the environment progresses. For the planning agents we use, the right delay is state-dependent, and naively planning how long to plan can paralyze the agent. We instead approach this setting by training a lightweight gating policy on top of a planner to select state-dependent planning budgets. Across real-time Pac-Man, Tetris, Snake, Speed Hex, and Speed Go, our gating policy outperforms fixed-budget and heuristic baselines, and transfers to a real-time setup where the environment and agent run on two different GPUs.

Reinforcement learning (RL) for locomotion frequently converges to locally optimal but undeployable behaviors, such as vibrating limbs or scooting on the torso, that maximize return without producing a usable gait. We present MPC-Injection, a low-overhead method that steers RL toward a designer-preferred behavior by inserting transitions generated in the same environment by a model predictive controller (MPC). Unlike reward shaping, MPC-Injection does not require redesigning the task reward, and unlike adversarial imitation learning, it adds no discriminator, no kinematic retargeting, and no auxiliary objective. We analyze how the injected transitions bias the learning, allowing the policy to converge to behaviors that pure RL may fail to reach under simple reward functions. On a 2D walker in simulation and with sim-to-real evaluation on a Go2 quadruped, we show that MPC-Injection produces gaits qualitatively comparable to those of reward shaping and adversarial motion priors. We also show that MPC-Injection can complete a barrel roll that pure RL fails to achieve under the same simple reward and can select between trotting and bounding gaits only through changing the injected MPC data.

Reinforcement learning in large or sparse-reward environments suffers from slow temporal-difference reward propagation, as value information spreads only locally across the state space. We propose Mesh-RL, a spatial domain-decomposition framework inspired by the finite element method and domain decomposition theory, which partitions the environment into overlapping subgrids and enforces boundary-consistent temporal-difference updates. Such an approach enables localized learning while ensuring globally coherent value propagation. Unlike hierarchical or model-based approaches, Mesh-RL accelerates long-range credit assignment without modifying the reward function, Bellman operator, or introducing explicit planning mechanisms. We evaluate Mesh-RL on hazard-dense grid-world environments with varying geometries and mesh resolutions. Across Q-learning, SARSA, and Dyna-Q, Mesh-RL consistently improves convergence speed, cumulative reward, and learning stability. Higher mesh resolutions sustain exploration, prevent premature convergence, and substantially accelerate value propagation to distant states. While Dyna-Q already benefits from internal planning, it still achieves additional gains under structured decomposition. Overall, Mesh-RL introduces a principled spatial domain-decomposition mechanism for accelerating temporal-difference learning. Our framework bridges finite element method-inspired boundary-consistency techniques from scientific computing with reinforcement learning to improve sample efficiency in sparse-reward environments. We will release source code of the study.

Vulnerability detection in C/C++ software remains a major security challenge due to code complexity, manual memory management, and the limitations of traditional static analysis. Reinforcement Learning (RL) has emerged as a promising approach, particularly for fuzzing, test generation, program exploration, and, more recently, vulnerability detection and localization. Following PRISMA 2020 guidelines, this work reviews RL techniques for software vulnerability analysis, focusing on C/C++ source code and static analysis. We identified 21 primary studies published between 2015 and 2026 from major scientific databases and complementary searches. We analyze the addressed tasks, algorithms, state-action-reward-environment formulations, code representations, datasets, and evaluation metrics. Results show that 15 studies focus on fuzzing and guided exploration, only 3 on direct vulnerability detection, and just 1 on statement-level localization. Moreover, statically extracted structural representations such as Control Flow Graphs (CFGs) and Abstract Syntax Trees (ASTs) are rarely used as agent states, and benchmarks lack comparability. We propose a task- and formulation-oriented taxonomy and identify a key research gap: the absence of RL agents that use source-code CFGs as states to detect and localize vulnerable nodes.

Reinforcement Learning (RL) has enabled LLMs to excel in objective reasoning tasks such as mathematics and code generation. However, applying RL to open-ended tasks, such as creative writing, remains challenging because LLM-as-a-judge reward models often exhibit stylistic biases and positional inconsistencies, leading to unstable supervision. To address this, we propose OPERA (Objective Perplexity-based Reflective Alignment), which replaces unreliable external judges with intrinsic rewards derived from perplexity dynamics. Specifically, we derive an intrinsic reward signal from perplexity dynamics, quantifying uncertainty reduction at critical reflective states. During the cold-start phase, we introduce a data synthesis method that leverages carefully designed guiding words to generate diverse reasoning traces, along with perplexity-prioritized rollouts that utilize internal log-probabilities to identify logically consistent reasoning branches. This pipeline yields a large-scale dataset comprising 20,000 high-quality reasoning trajectories. Empirical evaluations consistently demonstrate the scalability and efficacy of our approach in alignment for open-ended tasks. Implementing OPERA on Qwen3-8B establishes a new state-of-the-art among open-source models, achieving parity with or surpassing proprietary models like Gemini2.5 and MiniMax-M2.5 in some open-ended tasks. The code is available at https://github.com/pangpang-xuan/OPERA.

We study optimal-policy geometry in structured Markov decision processes. While approximate dynamic programming and reinforcement learning typically approximate high-dimensional value functions, we show that optimal policies induce simpler decision tessellations. We propose boundary-based policy approximations that learn policy regions directly. A policy-loss decomposition links performance degradation to action margins and explains why errors concentrate near indifference boundaries. Inventory control and queue admission experiments show lower policy error, smaller value gaps, faster error decay, and stability than reinforcement learning baselines.

Living systems navigate environments using noisy and incomplete sensory signals. In unicellular algae, phototaxis is often modeled as a mechanistic run--tumble process driven by stimulus--response rules. However, such descriptions overlook how organisms actively sample their environment to reduce sensory ambiguity. From a minimal cognition perspective, we reframe this navigation as a subjective, information-driven sensorimotor process. To this end, we propose a framework linking a Partially Observable Markov Decision Process (POMDP) with biochemical reaction dynamics. Environmental variables are hidden, while the cell updates a minimal internal state from each observation through a memoryless Bayesian step. These internal dynamics balance orienting toward light with exploratory reorientation and can be implemented through Chemical-Reaction-Network Ordinary Differential Equations (CRN--ODEs). Our model includes a biophysical observation process for photoreception and a chemically computable polynomial bound on information gain. Using Inverse Reinforcement Learning (IRL) on 30 experimentally recorded Chlamydomonas trajectories, we infer the behavioral objective consistent with observed phototactic motion and benchmark the resulting dynamics with standard Stochastic Simulation Algorithm (SSA) baselines. Our model reproduces the empirical alignment-to-light distribution, comparable to objective SSA baselines on this dataset. Within this framework, run--tumble alternation emerges as an information-acquisition strategy: tumbling reorients the cell to sample new sensory configurations and resolve sensor ambiguity, demonstrating how intracellular biochemical networks can support adaptive information-seeking behavior in cellular navigation.