Related papers: Failure Modes of Maximum Entropy RLHF

Bootstrapping large language models (LLMs) through preference-based policy optimization offers a promising direction for aligning model behavior with human preferences without relying on extensive manual annotations. In this work, we…

Offline preference optimization offers a simpler and more stable alternative to RLHF for aligning language models. However, their effectiveness is critically dependent on ranking accuracy, a metric where further gains are highly impactful.…

Rerankers play a pivotal role in refining retrieval results for Retrieval-Augmented Generation. However, current reranking models are typically optimized on static human annotated relevance labels in isolation, decoupled from the downstream…

Existing studies on constrained reinforcement learning (RL) may obtain a well-performing policy in the training environment. However, when deployed in a real environment, it may easily violate constraints that were originally satisfied…

Score-function based methods for policy learning, such as REINFORCE and PPO, have delivered strong results in game-playing and robotics, yet their high variance often undermines training stability. Using pathwise policy gradients, i.e.…

Self-supervised reinforcement learning (RL) presents a promising approach for enhancing the reasoning capabilities of Large Language Models (LLMs) without reliance on expensive human-annotated data. However, we find that existing methods…

Reinforcement learning (RL) has been successfully applied to solve the problem of finding obstacle-free paths for autonomous agents operating in stochastic and uncertain environments. However, when the underlying stochastic dynamics of the…

Reinforcement learning (RL) has become a cornerstone for fine-tuning Large Language Models (LLMs), with Proximal Policy Optimization (PPO) serving as the de facto standard algorithm. Despite its ubiquity, we argue that the core ratio…

Group Relative Policy Optimization (GRPO) has been a key driver of recent progress in reinforcement learning with verifiable rewards (RLVR) for large language models, but it is typically trained in a low-staleness, near-on-policy regime…

On-policy reinforcement learning methods, like Trust Region Policy Optimization (TRPO) and Proximal Policy Optimization (PPO), often demand extensive data per update, leading to sample inefficiency. This paper introduces Reflective Policy…

Direct Preference Optimization (DPO) has become the de facto standard for offline preference alignment of large language models, but its reliance on a reference policy introduces a critical tension. DPO weighs each update relative to a…

Reinforcement learning with human feedback (RLHF) has become the dominant method to align large models to user preferences. Unlike fine-tuning, for which there are many studies regarding training data memorization, it is not clear how…

Offline reinforcement learning (RL) aims to find an optimal policy for Markov decision processes (MDPs) using a pre-collected dataset. In this work, we revisit the linear programming (LP) reformulation of Markov decision processes for…

Model-based reinforcement learning approaches carry the promise of being data efficient. However, due to challenges in learning dynamics models that sufficiently match the real-world dynamics, they struggle to achieve the same asymptotic…

Reinforcement learning (RL) has become a key approach for enhancing reasoning in large language models (LLMs), yet scalable training is often hindered by the rapid collapse of policy entropy, which leads to premature convergence and…

Recent advancements in post-training methodologies for large language models (LLMs) have highlighted reinforcement learning (RL) as a critical component for enhancing reasoning. However, the substantial computational costs associated with…

This effort is focused on examining the behavior of reinforcement learning systems in personalization environments and detailing the differences in policy entropy associated with the type of learning algorithm utilized. We demonstrate that…

Training reinforcement learning policies using environment interaction data collected from varying policies or dynamics presents a fundamental challenge. Existing works often overlook the distribution discrepancies induced by policy or…

Large language models (LLMs) benefit substantially from supervised fine-tuning (SFT) and reinforcement learning with verifiable rewards (RLVR) in reasoning tasks. However, these recipes perform poorly in instruction-based molecular…

Despite the efficacy of Direct Preference Optimization (DPO) in aligning Large Language Models (LLMs), reward hacking remains a pivotal challenge. This issue emerges when LLMs excessively reduce the probability of rejected completions to…