Quantile Advantage Estimation for Entropy-Safe Reasoning

• URL: http://arxiv.org/abs/2509.22611v1
• Date: Fri, 26 Sep 2025 17:37:52 GMT
• Title: Quantile Advantage Estimation for Entropy-Safe Reasoning
• Authors: Junkang Wu, Kexin Huang, Jiancan Wu, An Zhang, Xiang Wang, Xiangnan He,
• Abstract summary: Reinforcement Learning with Verifiable Rewards (RLVR) strengthens LLM reasoning, but training often oscillates between entropy collapse and entropy explosion<n>We trace both hazards to the mean baseline used in value-free RL, which improperly penalizes negative-advantage samples under reward outliers.<n>We propose Quantile Advantage Estimation (QAE), replacing the mean with a group-wise K-quantile baseline.
• Score: 44.192277495613695
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) strengthens LLM reasoning, but training often oscillates between {entropy collapse} and {entropy explosion}. We trace both hazards to the mean baseline used in value-free RL (e.g., GRPO and DAPO), which improperly penalizes negative-advantage samples under reward outliers. We propose {Quantile Advantage Estimation} (QAE), replacing the mean with a group-wise K-quantile baseline. QAE induces a response-level, two-regime gate: on hard queries (p <= 1 - K) it reinforces rare successes, while on easy queries (p > 1 - K) it targets remaining failures. Under first-order softmax updates, we prove {two-sided entropy safety}, giving lower and upper bounds on one-step entropy change that curb explosion and prevent collapse. Empirically, this minimal modification stabilizes entropy, sparsifies credit assignment (with tuned K, roughly 80% of responses receive zero advantage), and yields sustained pass@1 gains on Qwen3-8B/14B-Base across AIME 2024/2025 and AMC 2023. These results identify {baseline design} -- rather than token-level heuristics -- as the primary mechanism for scaling RLVR.

Related papers

• Overconfident Errors Need Stronger Correction: Asymmetric Confidence Penalties for Reinforcement Learning [17.384089089363382]
  We identify a root cause that existing methods overlook: the uniform penalization of errors.<n>Current approaches treat all incorrect rollouts within a group identically.<n>We propose the Asymmetric Confidence-aware Error Penalty (ACE)
  arXiv  Detail & Related papers  (2026-02-24T22:46:43Z)
• STAPO: Stabilizing Reinforcement Learning for LLMs by Silencing Rare Spurious Tokens [38.425692691443764]
  ExistingReinforcement Learning (RL) fine-tuning methods rely heavily on entropy regularization and reweighting to maintain stability.<n>In practice, they often suffer from late-stage performance collapse, leading to degraded reasoning quality and unstable training.<n>We find that training instability can be caused by a tiny fraction of tokens, approximately 0.01%, which we term spurious tokens.<n>We propose Spurious-Token-Aware Policy Optimization (STAPO), which promotes stable and effective large-scale model refinement.
  arXiv  Detail & Related papers  (2026-02-17T14:46:48Z)
• Outcome-Grounded Advantage Reshaping for Fine-Grained Credit Assignment in Mathematical Reasoning [60.00161035836637]
  Group Relative Policy Optimization has emerged as a promising critic-free reinforcement learning paradigm for reasoning tasks.<n>We introduce Outcome-grounded Advantage Reshaping (OAR), a fine-grained credit assignment mechanism that redistributes advantages based on how much each token influences the model's final answer.<n>OAR-G achieves comparable gains with negligible computational overhead, both significantly outperforming a strong GRPO baseline.
  arXiv  Detail & Related papers  (2026-01-12T10:48:02Z)
• Beyond High-Entropy Exploration: Correctness-Aware Low-Entropy Segment-Based Advantage Shaping for Reasoning LLMs [6.948242693954442]
  Reinforcement Learning with Verifiable Rewards (RLVR) has become a central approach for improving the reasoning ability of large language models.<n>We propose LESS, a correctness-aware reinforcement framework that performs fine-grained advantage modulation over low-entropy segments.
  arXiv  Detail & Related papers  (2025-11-30T14:19:36Z)
• Mind Your Entropy: From Maximum Entropy to Trajectory Entropy-Constrained RL [56.085103402298905]
  We propose a trajectory entropy-constrained reinforcement learning (TECRL) framework to address these two challenges.<n>Within this framework, we first separately learn two Q-functions, one associated with reward and the other with entropy, ensuring clean and stable value targets unaffected by temperature updates.<n>We develop a practical off-policy algorithm, DSAC-E, by extending the state-of-the-art distributional soft actor-critic with three refinements.
  arXiv  Detail & Related papers  (2025-10-25T09:17:47Z)
• Rediscovering Entropy Regularization: Adaptive Coefficient Unlocks Its Potential for LLM Reinforcement Learning [55.59724323303857]
  We propose a framework that balances exploration and exploitation via three components: difficulty-aware coefficient allocation, initial-anchored target entropy, and dynamic global coefficient adjustment.<n>Experiments on multiple mathematical reasoning benchmarks show that AER consistently outperforms baselines, improving both reasoning accuracy and exploration capability.
  arXiv  Detail & Related papers  (2025-10-13T03:10:26Z)
• Reinforcement Learning with Verifiable yet Noisy Rewards under Imperfect Verifiers [90.50039419576807]
  Reinforcement Learning with Verifiable Rewards (RLVR) trains policies against automated verifiers to avoid costly human labeling.<n>To reduce vulnerability to verifier hacking, many RLVR systems collapse rewards to binary $0,1$ during training.<n>This choice carries a cost: it introduces textitfalse negatives (rejecting correct answers, FNs) and textitfalse positives (accepting incorrect ones, FPs)
  arXiv  Detail & Related papers  (2025-10-01T13:56:44Z)
• Clip-Low Increases Entropy and Clip-High Decreases Entropy in Reinforcement Learning of Large Language Models [29.822717720666134]
  We show that the clipping mechanism in PPO and GRPO induces biases on entropy.<n>With a more aggressive clip-low value, one can increase entropy, promote exploration, and ultimately prevent entropy collapse in RLVR training.
  arXiv  Detail & Related papers  (2025-09-30T11:33:15Z)
• The Choice of Divergence: A Neglected Key to Mitigating Diversity Collapse in Reinforcement Learning with Verifiable Reward [58.559544190947584]
  A central paradox in fine-tuning Large Language Models (LLMs) with Reinforcement Learning with Verifiable Reward (RLVR) is the frequent degradation of multi-attempt performance.<n>This is often accompanied by catastrophic forgetting, where models lose previously acquired skills.<n>We argue that standard RLVR objectives lack a crucial mechanism for knowledge retention.
  arXiv  Detail & Related papers  (2025-09-09T06:34:32Z)
• Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning [80.87085014818052]
  Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a powerful approach to enhancing the reasoning capabilities of Large Language Models (LLMs)<n>In this work, we undertake a pioneering exploration of RLVR through the novel perspective of token entropy patterns.<n>We observe that only a small fraction of tokens exhibit high entropy, and these tokens act as critical forks that steer the model toward diverse reasoning pathways.
  arXiv  Detail & Related papers  (2025-06-02T17:54:39Z)
• Stable Reinforcement Learning for Efficient Reasoning [2.838966689544288]
  GRPO-$lambda$ is an efficient and stabilized variant of GRPO.<n>It dynamically adjusts the reward strategy by monitoring the correctness ratio.<n>It improves average accuracy by 1.48% while reducing CoT sequence length by 47.3%.
  arXiv  Detail & Related papers  (2025-05-23T16:43:03Z)
• A Piecewise Lyapunov Analysis of Sub-quadratic SGD: Applications to Robust and Quantile Regression [22.917692982875025]
  We introduce a novel piecewise Lyapunov function that enables us to handle functions $f$ with only first-order differentiability.<n>We derive finite-time moment bounds under general diminishing stepsizes, as well as constant stepsizes.<n>Our results have wide applications, especially in online statistical methods.
  arXiv  Detail & Related papers  (2025-04-11T00:20:37Z)
• Tilted Quantile Gradient Updates for Quantile-Constrained Reinforcement Learning [12.721239079824622]
  We propose a safe reinforcement learning (RL) paradigm that enables a higher level of safety without any expectation-form approximations.<n>A tilted update strategy for quantile gradients is implemented to compensate the asymmetric distributional density.<n>Experiments demonstrate that the proposed model fully meets safety requirements (quantile constraints) while outperforming the state-of-the-art benchmarks with higher return.
  arXiv  Detail & Related papers  (2024-12-17T18:58:00Z)
• Understanding, Predicting and Better Resolving Q-Value Divergence in Offline-RL [86.0987896274354]
  We first identify a fundamental pattern, self-excitation, as the primary cause of Q-value estimation divergence in offline RL. We then propose a novel Self-Excite Eigenvalue Measure (SEEM) metric to measure the evolving property of Q-network at training. For the first time, our theory can reliably decide whether the training will diverge at an early stage.
  arXiv  Detail & Related papers  (2023-10-06T17:57:44Z)
• Approximation Schemes for ReLU Regression [80.33702497406632]
  We consider the fundamental problem of ReLU regression. The goal is to output the best fitting ReLU with respect to square loss given to draws from some unknown distribution.
  arXiv  Detail & Related papers  (2020-05-26T16:26:17Z)

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.