This paper studies the distribution of reinforcement learning (RL) adaptation across transformer layers in large language models and finds that training a single layer can recover most of the gains ob…

This paper synthesizes expert knowledge from a workshop to provide a comprehensive framework and best-practice guidelines for developing high-quality reinforcement learning environments for autonomous…

This paper proposes using offline reinforcement learning (RL) as an efficient alternative to online RL for post-training code-generating LLMs, demonstrating its effectiveness, especially for smaller m…

This paper explores approaches to generating behavioral diversity in reinforcement learning models to bridge the gap between simulation and biology.

This paper proposes a continual RL framework based on Continual Backpropagation for autonomous racing, able to train a generalistic policy on a set of tracks and fine-tune it within 15 minutes to outp…

This paper introduces a novel backdoor attack (ACB) against Reinforcement Learning with Verifiable Rewards (RLVR), demonstrating that poisoning the training data can implant a backdoor that significan…

ARES is a novel framework that systematically discovers and mitigates dual vulnerabilities in RLHF systems by simultaneously testing the core LLM and its Reward Model (RM) using structured adversarial…

The paper proposes ReuseRL, a method that improves agent generalization in Reinforcement Learning by enforcing structural compressibility of successful agent trajectories into reusable skills.

The paper introduces $ ext{RLR}^3$, a novel framework that extends verifiable rewards in Reinforcement Learning to handle partially verifiable, multi-criteria vision-language tasks by integrating robu…

The paper introduces Atomic Decomposition and Recombination (ADR), a novel framework that generates genuinely novel and challenging verifiable code tasks, significantly improving the scalability of Re…

IRDS introduces a novel data selection method that uses a verifier-coupled sparse autoencoder framework to efficiently select high-quality Reinforcement Learning with Verifiable Rewards (RLVR) trainin…

The paper proposes a Digital Twin-assisted Adaptive Multi-Agent Deep Reinforcement Learning framework to intelligently manage spectrum and resources in complex, dynamic Open-RAN 6G networks utilizing…

The paper introduces a verifier-fuzzing framework to detect and analyze failure modes in Reinforcement Learning with Verifiable Rewards (RLVR) where bugs in the reward verifier can be exploited by the…

This paper theoretically justifies the strong performance of linear recurrent neural networks as memory units in partially observable reinforcement learning by constructing specific linear filters tha…

The paper introduces OpenWebRL, an open framework that enables training visual web agents using online multi-turn Reinforcement Learning directly on live websites, achieving state-of-the-art performan…

The paper proposes a Hierarchical Reinforcement Learning framework with two levels for handling high-level strategic planning and low-level continuous-control using Soft Actor-Critic and entropy-regul…

The paper introduces Group Prioritized Off-Policy Optimization (POPO), a novel framework that efficiently accelerates RL finetuning for LLM reasoning by leveraging effective off-policy training batche…