NeurIPS 2026 Papers

Generative policies based on expressive model classes, such as diffusion and flow matching, are well-suited to complex control problems with highly multimodal action distributions. Their expressivity, however, comes at a significant inference cost: generating each action typically requires simulating many steps of the generative process, compounding latency across sequential decision-making rollouts. We introduce flow map policies, a novel class of generative policies designed for fast action generation by learning to take arbitrary-size jumps---including one-step jumps---across the generative dynamics of existing flow-based policies. We instantiate flow map policies for offline-to-online reinforcement learning and formulate online adaptation as a trust-region optimization problem that improves the critic's $Q$-value while remaining close to the offline policy. We theoretically derive Flow Map $Q$-Guidance Training (FMQ), a principled closed-form learning target that is optimal for adapting offline flow map policies under a critic-guided trust-region constraint. We further introduce $Q$-Guided Beam Search (QGBS), a stochastic flow-map sampler that combines renoising with beam search to enable iterative inference-time refinement. Across $12$ challenging robotic manipulation and locomotion tasks from OGBench and RoboMimic, FMQ achieves state-of-the-art performance in offline-to-online RL, outperforming the previous one-step policy MVP by a relative improvement of $21.3$% on the average success rate.

Reinforcement learning with verifiable rewards (RLVR) provides a promising pathway for continuously advancing GUI agents, yet existing reward modeling paradigms face complementary limitations. Rule-based methods suffer from poor scalability and cannot handle open-ended tasks. LLM-as-a-Judge methods enable scalable trajectory verification but remain passive and are constrained by partial state observability, since key evidence often resides in latent environment states beyond the trajectory. Recent active environment interaction methods mitigate observability issues but tend to over-rely on probing while under-utilizing direct trajectory evidence, leading to verification inefficiency. To address these challenges, we advocate a trajectory-grounded interactive verification paradigm. We introduce VAGEN, a framework that employs a tool-augmented verifier agent governed by a Progressive Verification Mechanism, which follows a surface-to-latent and cheap-to-expensive design philosophy to extract trajectory evidence and probe environment states in a proactive end-to-end manner. Experiments on OSWorld-Verified and AndroidWorld benchmarks demonstrate that VAGEN significantly improves evaluation accuracy with a favorable performance-efficiency trade-off.

LLM-based multi-agent systems are increasingly deployed on long-horizon tasks, but a single decisive error is often accepted by downstream agents and cascades into trajectory-level failure. Existing work frames this as \emph{post-hoc failure attribution}, diagnosing the responsible agent and step after the trajectory has ended. However, this paradigm forfeits any opportunity to intervene while trajectory is still unfolding. In this work, we introduce AgentForesight, a framework that reframes this problem as online auditing: at each step of an unfolding trajectory, an auditor observes only the current prefix and must either continue the run or alarm at the earliest decisive error, without access to future steps. To this end, we curate AFTraj-2K, a corpus of agentic trajectories across Coding, Math, and Agentic domains, in which safe trajectories are retained under a strict curation pipeline and unsafe trajectories are annotated at the step of their decisive error via consensus among multiple LLM judges. Built on that, we develop AgentForesight-7B, a compact online auditor trained with a coarse-to-fine reinforcement learning recipe that first equips it with a risk-anticipation prior at the failure boundary on adjacent safe/unsafe prefix pairs, then sharpens this prior into precise step-level localization under a three-axis reward jointly targeting the what, where, and who of an audit verdict. Across AFTraj-2K and an external Who\&When benchmark, AgentForesight-7B outperforms leading proprietary models, including GPT-4.1 and DeepSeek-V4-Pro, achieving up to +19.9\% performance gain and 3$\times$ lower step localization error, opening the loop from post-hoc failure detection to enabling deployment-time intervention.

Open-Vocabulary Multimodal Emotion Recognition (OV-MER) aims to predict emotions without being constrained by predefined label spaces, thereby enabling fine-grained emotion understanding. Unlike traditional discriminative methods, OV-MER leverages generative models to capture the full spectrum of emotions and employs emotion wheels (EWs) for metric calculation. Previous approaches primarily rely on token-level loss during training. However, this objective is misaligned with the metrics used in OV-MER, and these metrics cannot be directly optimized via gradient backpropagation. To address this limitation, we turn our attention to reinforcement learning, as this strategy can optimize non-differentiable objectives. We term this framework **AffectGPT-RL**. Furthermore, we conduct extensive experiments to elucidate the role of reinforcement learning in this task, revealing the necessity of the reasoning process, the impact of different rewards, and the generalizability to other emotion tasks such as sentiment analysis and basic emotion recognition. Experimental results demonstrate that AffectGPT-RL yields significant performance improvements on OV-MER. Beyond this task, we also achieve remarkable performance gains on basic emotion recognition, attaining state-of-the-art results on MER-UniBench. To the best of our knowledge, this is the pioneering work exploring the role of reinforcement learning in OV-MER, providing valuable guidance for subsequent researchers. Our code is provided in the supplementary material and will be released to facilitate future research.

We propose FAR-SIGN (Fully Asynchronous Robust optimization via SIGNed directional projections) for adversary-resilient learning in parameter-server--worker systems. FAR-SIGN achieves robustness through sign-based updates along carefully designed directions and mitigates the resulting bias via a two-timescale mechanism. It admits both first-order and zeroth-order implementations and enables fully asynchronous execution without requiring a private reference dataset at the server. We establish almost-sure convergence of FAR-SIGN to the set of stationary points for smooth, nonconvex objectives. Moreover, we prove the near-optimal rate of $O(n^{-1/4+\epsilon})$ in the first-order setting and the standard $O(n^{-1/6+\epsilon})$ in the zeroth-order setting, where $n$ is the iteration count and $\epsilon>0$ can be chosen arbitrarily small. Experiments on MNIST show that FAR-SIGN outperforms robust aggregation-based methods in both accuracy and wall-clock time.

Temporally extended exploration via graph Laplacian-based options is a promising approach to sparse-reward reinforcement learning (RL), but existing methods either do not explicitly target novelty or fail to scale to pixel-based domains under function approximation. Novel Exploration via Orthogonality (NEO) addresses the first issue by constructing options that navigate from highly visited regions toward less visited ones, yet prior results were limited to settings where exact eigenvectors can be computed. We present a scalable extension of NEO to pixel-based domains, built on three contributions. First, we use a novelty-weighted continuous Laplacian graph-drawing objective, which enables RL with continuous observations. Second, we embed the resulting eigen-potential options within a hierarchical reinforcement learning framework, enabling coherent temporally extended behavior. Third, we observe that learned eigen-potential rewards are directional but locally unreliable under online approximation; we therefore augment each option reward with a novelty bonus, a novel design idea that proves essential for stabilizing option learning while preserving novelty-directed exploration. Together, these contributions yield stronger and more persistent exploration, enabling longer option rollouts and better access to hard-to-reach novel states. Empirically, our method significantly outperforms both the prior scalable Laplacian-option baseline and a direct extension of NEO on sparse-reward benchmarks under a fixed budget. On Montezuma's Revenge, our best variant achieves approximately 1.8x higher return than both baselines. On Venture, both baselines yield returns near zero, whereas our method achieves a return of 1135. Across seven hard ProcGen games, our method achieves approximately 3.5x and 5.6x higher aggregate normalized return than the two baselines, respectively.

Cooperative multi-agent reinforcement learning (MARL) often fails under exogenous variability such as shifts in opponents or environment regimes that cannot be controlled by the team but alter the transition dynamics. We formalize this challenge as Exogenous Dec-POMDP (ED-POMDP), which decomposes the global state into endogenous variables controllable by the team and exogenous variables beyond its control. This formulation exposes delayed influence: While actions do not directly determine exogenous transitions, they can shape them through induced changes in endogenous state. Based on this, we propose LEICA, a CTDE-compatible algorithm that learns history-conditioned endogenous and exogenous context representations and shapes policy updates using influence-weighted intrinsic rewards. Across SMAX benchmarks with opponent strategy shifts, LEICA consistently improves both training performance and generalization to unseen opponents' strategies over existing baselines, supporting the usefulness of influence-weighted shaping under exogenous train-test regime shifts.

Evolution strategies (ES) have recently emerged as a promising alternative to gradient-based reinforcement learning for fine-tuning large language models, but classical zeroth-order optimization theory does not systematically account for the empirical observations driving this success. We formulate the continuous-time limit of Z-score-normalized ES as a unified stochastic differential equation and identify three sequential dynamical regimes characterized by the dominant gradient contribution to the population reward variance; under a basin-local low-rank Hessian assumption, this framework systematically explains (1) the information-geometric optimality of the Z-score heuristic, (2) the dimensionality paradox of sample-efficient ES with populations far smaller than the parameter dimension, (3) the same-task rise-then-decay of the training reward, and (4) catastrophic forgetting of pre-training capabilities. Synthetic experiments and direct verification on Qwen2.5-{0.5B,3B,7B}-Instruct confirm these predictions, matching the parameter drift reported by Abdi et al. (2026) within $\sim 1.5\\%$.

Modern deep reinforcement learning algorithms often store past sensory observations in a buffer for later replay and learning. However, this process is biologically implausible; animals do not store raw sensory observations. Instead, they store their imperfect understanding of the world and the events in it. An alternative to storing all past sensory observations is to learn directly from the observations as they come in and store the seemingly important parts of the incoming data. Eligibility trace algorithms in reinforcement learning (RL) sweep back updates for multiple steps, improving efficiency and providing update rules that do not rely on storing large data buffers. These algorithms were common and effectively used in the linear setting, but have not been widely adopted in the deep RL setting. Many questions remain open to make these algorithms more effective, even as basic as how to incorporate vector step-sizes. We first highlight how adaptive vector step-sizes need to be incorporated into the eligibility trace and then introduce a new vector step-size algorithm that is more amenable to the forward-backward view equivalence needed to derive an update with eligibility traces. We show that our new algorithms improve over other streaming algorithms across Mujoco and MinAtar environments, and are comparable to methods with replay buffers such as PPO.

We present the Adaptive Scheduling Pipeline (ASP), a bounded-lag asynchronous reinforcement learning (RL) architecture for post-training large language models (LLMs). Rather than only disaggregating generation and training, the ASP combines two design ideas. First, a queue-aware scheduling policy that jointly reduces three sources of rollout underutilization —padding bubbles from intra-minibatch length heterogeneity, skewness bubbles from long-tail outputs, and sorting bubbles from queue instability induced by length-sorted dispatching—through a single offline preprocessing pass. This enables efficient multi-instance rollout that overcomes the early scaling plateau of single-instance inference while maintaining stable queue dynamics. Second, a bounded-lag asynchronous architecture decouples rollout from training while keeping policy lag observable and small enough to preserve stable learning in our experiments. We also provide an explicit weight-flow path that converts trainer snapshots into inference-ready weights without re-coupling generation and training. Experiments show that ASP improves end-to-end throughput by up to $3.22\times$, reduces iteration time by up to 76\%, scales more robustly with additional generation resources than the MindSpeed-RL and verl baselines, and shows no visible convergence degradation. ASP has also been used in an internal telecom-domain RL post-training workflow, reducing the reported training period from more than three months to about two months and producing a domain-enhanced model for knowledge answering and tool calling.

Reinforcement learning and contextual bandit algorithms have become increasingly common in sequential decision-making applications. When these methods are deployed in high-stakes domains, there is growing interest not only in learning effective policies, but also in conducting statistical inference for quantities learned under adaptive data collection. However, classical procedures applied naively in these settings can fail: even when estimators are unbiased, their variance becomes path-dependent and as a result may not be asymptotically normal. A growing literature has emerged to ameliorate this problem, but solutions tend to be problem specific and often rely on correct specification of a working model. In this work, we develop a unified framework for constructing asymptotically valid confidence intervals to cover smooth functionals of nonparametric M-estimands under adaptive sampling. Under Neyman orthogonality, we provide two novel methods for performing inference: (1) a self-normalized statistic based on the realized quadratic variation of the influence function and (2) a statistic using a plug-in estimate of the conditional variance based on reweighted influence function increments. Our results allow for flexible nonparametric estimation of nuisance parameters and remain valid under model misspecification. Our theory is supported by a simulation study for a dynamic pricing application which demonstrates that this method can produce asymptotically valid confidence intervals where standard methods fail.

Cooperative multi-agent reinforcement learning (MARL) has long faced scalability challenges due to the exponential growth of the state-action space as the number of agents increases. Existing methods typically enhance scalability by filtering out communications with low-relevance agents. However, such filtering often relies on global state and fixed graph distribution, limiting adaptability in large-scale, communication-constrained environments. In this paper, we propose a scalable MARL method, called Adaptive Communication Range PPO (ACR-PPO), that decomposes the decision-making under communication budget constraints as a sequential process: a communication policy first selects each agent’s communication range within a given budget, followed by a behavior policy that conditions actions on the resulting neighborhood observations. More importantly, we provide a theoretical guarantee of monotonic performance improvement under communication budget constraints. Experiments across diverse scenarios demonstrate that ACR-PPO preserves policy performance while significantly reducing communication costs through adaptive range control.

Offline goal-conditioned reinforcement learning (GCRL) aims to learn a single policy that adapts its behavior to arbitrary goals from a fixed dataset, making injection of the goal information into the policy a central design question. While recent works increasingly leverage expressive conditional generative modeling architectures, which have demonstrated remarkable success in computer vision, architectural choices for goal conditioning in offline GCRL remain under-explored. We identify that goals in GCRL differ from typical conditioning signals in two ways: they are often noisy or redundant, and they are consumed within a time-sequential episode where the relevant context of a fixed goal shifts as the agent's state evolves. Motivated by these observations, we propose State-Anchored Goal Conditioning (SAGC), a modulation-based goal conditioning scheme that derives its scale and shift parameters by jointly processing state and goal representations in a learned latent space. We further introduce AdaptFlow, an end-to-end trainable offline GCRL architecture that integrates conditional flow matching with SAGC. AdaptFlow outperforms matches all baselines on 23 out of 27 OGBench tasks.

Human players in cooperative games often rely on conventions: reusable patterns of play that coordinate well when used by the whole team, but can fail otherwise. Evaluating agents in this setting is difficult because human demonstration datasets are usually unlabeled: trajectories do not identify which player, group, or convention produced them. As a result, standard proxy-based evaluations can miss important population diversity. In particular, proxy bots produced by regularised self-play around a pooled behavioural cloning policy can remain concentrated around a single dominant mode of play. We introduce a pipeline for recovering and evaluating this hidden convention diversity from unlabeled demonstrations. Our method first learns latent convention structure from the dataset, then selects a compact set of distinct but plausible conventions, and finally reconstructs each selected convention into a high-performing playable proxy using KL-regularised PPO. A key practical challenge is that KL-regularised PPO is sensitive to the KL penalty, which is often tuned using online interaction. We provide an empirical offline calibration procedure for choosing this KL penalty in the PPO+KL reconstruction settings studied here. On Hanabi, we find that the BC-derived proxy suite is highly concentrated, and that the relative ordering of agents changes when evaluation uses the latent proxy suite instead.

Policy Dual Averaging (PDA) offers a principled Policy Mirror Descent (PMD) framework that more naturally admits value function approximation than standard PMD, enabling the use of approximate advantage (or Q-) functions while retaining strong convergence guarantees. However, applying PDA in continuous state and action spaces remains computationally challenging, since action selection involves solving an optimization sub-problem at each decision step. In this paper, we propose actor-accelerated PDA, which uses a learned policy network to approximate the solution of the optimization sub-problems, significantly reducing runtime. We provide a theoretical analysis that quantifies how actor approximation error impacts the convergence of PDA under suitable assumptions. We then evaluate its performance on several benchmarks in robotics, control, and operations research problems. Actor-accelerated PDA achieves superior performance compared to popular on-policy baselines such as Proximal Policy Optimization (PPO). Overall, our results take a significant step toward bridging the gap between the theoretical advantages of PDA and its practical deployment in continuous-action problems with function approximation.

We investigate the problem of scaling safe reinforcement learning to massively parallel training regimes, where a large number of environments are executed with short rollout horizons. Although this setting improves data throughput and wall-clock efficiency, it creates a mismatch with constrained Markov decision processes, whose safety constraints are defined over full-episode cost returns. We address this mismatch by analyzing staggered environment resets as a phase-mixture distribution over short trajectory segments. This analysis characterizes the bias induced by phase coverage and motivates a phase-aggregated cost estimator that reconstructs episode-level cost estimates from staggered short rollout batches. We further incorporate finite-sample uncertainty and distribution mismatch into a constraint-tightening framework, yielding a scalable Safe RL method compatible with on-policy optimization.Using an MJX-based implementation of Safety Gym, our experiments show that the proposed method achieves performance comparable to non-massively parallel implementations while substantially reducing wall-clock training time

Enforcing state-wise safety constraints is critical for the application of reinforcement learning (RL) in real-world problems, such as autonomous driving and robot manipulation. However, existing safe RL methods often enforce state-wise constraints only in expectation, while hard state-wise guarantees typically require strong dynamics assumptions. The former can still allow rare but severe violations, while the latter is impractical in model-free stochastic systems. We instead target high-probability control of the maximum instantaneous cost along a trajectory. To accomplish this goal, we propose Absolute State-wise Constrained Policy Optimization (ASCPO), a model-free policy search algorithm whose exact update provides a distribution-free high-probability certificate for state-wise constraint satisfaction in stochastic systems. The guarantee holds under finite variance and does not require Gaussian, unimodal, or light-tailed cost distributions. We demonstrate the effectiveness of our approach by training neural network policies for extensive robot locomotion tasks, where the agent must adhere to various state-wise safety constraints. Our results show that ASCPO most consistently reduces state-wise safety violations while maintaining competitive reward across challenging continuous control tasks.

Complementarity Determining Regions (CDRs) are critical segments of an antibody that facilitate binding to specific antigens. Current computational methods for CDR design rely on reconstruction losses and do not jointly optimize binding energy, a crucial metric of antibody efficacy. Rather, binding energy optimization is performed via computationally expensive Online Reinforcement Learning (RL) pipelines that rely heavily on unreliable binding-energy estimators. In this paper, we propose AbFlowNet, a novel generative framework that integrates GFlowNet with Diffusion models. By framing each diffusion step as a state in the GFlowNet framework, AbFlowNet jointly optimizes standard diffusion losses and binding energy by directly incorporating energy signals into the training process, thereby unifying diffusion and reward optimization in a single procedure. Experimental results show that AbFlowNet outperforms the base diffusion model by 3.06% in amino acid recovery, 20.40% in geometric reconstruction (RMSD), and 3.60% in binding energy improvement ratio. ABFlowNet also decreases Top-1 total energy and binding energy errors by 24.8% and 38.1%, which is competitive with AbDPO's expensive online RL optimization without pseudo-labeling the test dataset or requiring computationally expensive synthetic CDR sampling.

Professional image retouching is a sequential and reasoning-driven process. Photographers first analyze scene semantics, lighting structure, and depth cues to design a retouching plan, and then progressively apply global tone grading and mask-guided local corrections based on the evolving visual outcome. Recent MLLM-empowered retouching agents either predict global parameters or respond to user-provided retouching instructions. Although showing impressive results, they lack the capacity for autonomous sequential reasoning and fall short of fine-grained local control. We propose **AURA**, an **AU**tonomous **R**etouching **A**gent that emulates the workflow of expert photographers. Given an input photograph, AURA autonomously assesses the scene and proceeds progressively. At each step, it reasons about what to retouch, executes the adjustment, and feeds the result back as visual context for next-step reasoning, forming a fine-grained visual chain-of-thought that evolves with the image. To support AURA, we contribute the first expert-annotated long-horizon retouching trajectory dataset, where expert photographers edit the input image from scratch with deliberate artistic intent. Based on this dataset, we train AURA in two stages: supervised fine-tuning to learn the retouching trajectory, followed by GRPO-retouching, an agentic reinforcement learning stage that optimizes for perceptual quality and tool accuracy. Our experiments demonstrate that AURA consistently outperforms existing methods in both quantitative metrics and user studies, enabling one-click professional retouching and democratizing masterpiece previously accessible only to skilled experts.