Predictive Divergence Masks for LLM RL

• URL: http://arxiv.org/abs/2607.10848v1
• Date: Sun, 12 Jul 2026 17:20:44 GMT
• Title: Predictive Divergence Masks for LLM RL
• Abstract summary: Reinforcement learning for large language models (LLMs) typically relies on trust-region masks to stabilize off-policy updates.<n>Recent work DPPO improves the proximity criterion by replacing PPO's ratio-based test with a probability divergence between the behavior and training policies.<n>We propose the predictive divergence mask, which asks whether the next policy-gradient step will increase or decrease the same divergence used by the trust region.
• Score: 44.851813004352046
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Reinforcement learning for large language models (LLMs) typically relies on trust-region masks to stabilize off-policy updates. The dominant PPO-style approach uses the sampled-token importance ratio for two criteria: a proximity criterion, which asks whether the policy has moved too far from the behavior policy, and a direction criterion, which asks whether the update pushes it farther away. Recent work DPPO improves the proximity criterion by replacing PPO's ratio-based test with a probability divergence between the behavior and training policies. However, its direction criterion is still inherited from PPO. A token can be masked only when the sampled-token importance ratio moves away from one. We observe that this ratio-based direction criterion is a single-sample proxy that can disagree in sign with the change of the divergence that defines the proximity criterion. We therefore propose the predictive divergence mask, which asks whether the next policy-gradient step will increase or decrease the same divergence used by the trust region. For the discrete softmax policies used in LLM RL, we derive this prediction in closed form. Because production rollout engines expose only a truncated (top-K) view of the vocabulary, we develop two lightweight top-$K$ estimators for this prediction. Detailed analysis shows the divergence-based direction is better aligned with the realized change of the divergence than the sampled ratio, and the resulting masks improve RL training across model scales and precision settings.

Related papers

• Flow-DPPO: Divergence Proximal Policy Optimization for Flow Matching Models [61.74572554094633]
  We argue that ratio clipping is structurally ill-suited for flow models.<n>We propose Flow-DPPO, which replaces ratio clipping with a divergence proximal constraint.<n> Experiments show that Flow-DPPO achieves higher rewards with better KL-proximal efficiency.
  arXiv  Detail & Related papers  (2026-06-09T15:59:57Z)
• Rethinking the Divergence Regularization in LLM RL [56.952725733148746]
  Reinforcement learning (RL) has become a key component of post-training large language models (LLMs)<n>We propose Divergence Regularized Policy Optimization (DRPO), which replaces the hard mask with a smooth advantage-weighted quadratic regularizer on policy shift.<n>DRPO preserves the same trust-region geometry as DPPO while inducing bounded, continuous gradient weights.
  arXiv  Detail & Related papers  (2026-06-08T17:58:23Z)
• Rethinking Importance Sampling in LLM Policy Optimization: A Cumulative Token Perspective [22.848847562976633]
  Existing methods face a fundamental bias-variance dilemma: token-level IS ratios, as adopted by PPO and GRPO, introduce bias by ignoring prefix state distribution mismatch.<n>We propose CTPO, which combines the cumulative token IS ratio with position-adaptive clipping that scales log-space clip bounds according to the natural $sqrtt$ growth of the cumulative log-ratio.
  arXiv  Detail & Related papers  (2026-05-08T06:35:02Z)
• Improving Search Agent with One Line of Code [68.58667107354253]
  Tool-based Agentic Reinforcement Learning (TARL) has emerged as a promising paradigm for training search agents to interact with external tools.<n>We propose textbfSearch textbfAgent textbfPolicy textbfOptimization (textbfSAPO), which stabilizes training via a conditional token-level KL constraint.
  arXiv  Detail & Related papers  (2026-03-10T04:07:39Z)
• Rethinking the Trust Region in LLM Reinforcement Learning [72.25890308541334]
  Proximal Policy Optimization (PPO) serves as the de facto standard algorithm for Large Language Models (LLMs)<n>We propose Divergence Proximal Policy Optimization (DPPO), which substitutes clipping with a more principled constraint.<n>DPPO achieves superior training and efficiency compared to existing methods, offering a more robust foundation for RL-based fine-tuning.
  arXiv  Detail & Related papers  (2026-02-04T18:59:04Z)
• A Step Back: Prefix Importance Ratio Stabilizes Policy Optimization [58.116300485427764]
  Reinforcement learning post-training can elicit reasoning behaviors in large language models.<n> token-level correction often leads to unstable training dynamics when the degree of off-policyness is large.<n>We propose a simple yet effective objective, Minimum Prefix Ratio (MinPRO)
  arXiv  Detail & Related papers  (2026-01-30T08:47:19Z)
• Off-Policy Evaluation in Markov Decision Processes under Weak Distributional Overlap [3.351714665243138]
  We re-visit the task of off-policy evaluation in Markov decision processes (MDPs) under a weaker notion of distributional overlap.<n>We introduce a class of truncated doubly robust (TDR) estimators which we find to perform well in this setting.<n>We find that, in our experiments, appropriate truncation plays a major role in enabling accurate off-policy evaluation when strong distributional overlap does not hold.
  arXiv  Detail & Related papers  (2024-02-13T03:55:56Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.