OBLR-PO: A Theoretical Framework for Stable Reinforcement Learning

• URL: http://arxiv.org/abs/2511.23310v1
• Date: Fri, 28 Nov 2025 16:09:28 GMT
• Title: OBLR-PO: A Theoretical Framework for Stable Reinforcement Learning
• Authors: Zixun Huang, Jiayi Sheng, Zeyu Zheng,
• Abstract summary: We provide a unified theoretical framework that characterizes the statistical properties of commonly used policy-gradient estimators.<n>We derive an adaptive learning-rate schedule governed by the signal-to-noise ratio (SNR) of gradients.<n>We further show that the variance-optimal baseline is a gradient-weighted estimator, offering a new principle for variance reduction.
• Score: 12.77713716713937
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Existing reinforcement learning (RL)-based post-training methods for large language models have advanced rapidly, yet their design has largely been guided by heuristics rather than systematic theoretical principles. This gap limits our understanding of the properties of the gradient estimators and the associated optimization algorithms, thereby constraining opportunities to improve training stability and overall performance. In this work, we provide a unified theoretical framework that characterizes the statistical properties of commonly used policy-gradient estimators under mild assumptions. Our analysis establishes unbiasedness, derives exact variance expressions, and yields an optimization-loss upper bound that enables principled reasoning about learning dynamics. Building on these results, we prove convergence guarantees and derive an adaptive learning-rate schedule governed by the signal-to-noise ratio (SNR) of gradients. We further show that the variance-optimal baseline is a gradient-weighted estimator, offering a new principle for variance reduction and naturally enhancing stability beyond existing methods. These insights motivate Optimal Baseline and Learning-Rate Policy Optimization (OBLR-PO), an algorithm that jointly adapts learning rates and baselines in a theoretically grounded manner. Experiments on Qwen3-4B-Base and Qwen3-8B-Base demonstrate consistent gains over existing policy optimization methods, validating that our theoretical contributions translate into practical improvements in large-scale post-training.

Related papers

• Stabilizing Policy Optimization via Logits Convexity [59.242732612484474]
  We show that the convexity of the supervised fine-tuning loss with respect to model logits plays a key role in enabling stable training.<n>Motivated by this observation, we propose Logits Convex Optimization (LCO), a simple yet effective policy optimization framework.
  arXiv  Detail & Related papers  (2026-03-01T07:40:12Z)
• ODELoRA: Training Low-Rank Adaptation by Solving Ordinary Differential Equations [54.886931928255564]
  Low-rank adaptation (LoRA) has emerged as a widely adopted parameter-efficient fine-tuning method in deep transfer learning.<n>We propose a novel continuous-time optimization dynamic for LoRA factor matrices in the form of an ordinary differential equation (ODE)<n>We show that ODELoRA achieves stable feature learning, a property that is crucial for training deep neural networks at different scales of problem dimensionality.
  arXiv  Detail & Related papers  (2026-02-07T10:19:36Z)
• 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)
• Alternating Reinforcement Learning for Rubric-Based Reward Modeling in Non-Verifiable LLM Post-Training [29.56905427210088]
  gradient-ARM is a framework that jointly optimize a rubric generator and a judge using reinforcement learning from preference feedback.<n>We show that gradient-ARM achieves state-of-the-art performance among baselines on benchmarks and significantly improves downstream policy alignment in both offline and online reinforcement learning settings.
  arXiv  Detail & Related papers  (2026-02-02T00:50:53Z)
• Deep Unfolding: Recent Developments, Theory, and Design Guidelines [99.63555420898554]
  This article provides a tutorial-style overview of deep unfolding, a framework that transforms optimization algorithms into structured, trainable ML architectures.<n>We review the foundations of optimization for inference and for learning, introduce four representative design paradigms for deep unfolding, and discuss the distinctive training schemes that arise from their iterative nature.
  arXiv  Detail & Related papers  (2025-12-03T13:16:35Z)
• AMStraMGRAM: Adaptive Multi-cutoff Strategy Modification for ANaGRAM [6.515592049126884]
  We analyze the training dynamics of PINNs optimized with ANaGRAM.<n>We propose a multi-cutoff adaptation strategy that further enhances ANaGRAM's performance.
  arXiv  Detail & Related papers  (2025-10-14T09:10:42Z)
• Stabilizing Policy Gradients for Sample-Efficient Reinforcement Learning in LLM Reasoning [77.92320830700797]
  Reinforcement Learning has played a central role in enabling reasoning capabilities of Large Language Models.<n>We propose a tractable computational framework that tracks and leverages curvature information during policy updates.<n>The algorithm, Curvature-Aware Policy Optimization (CAPO), identifies samples that contribute to unstable updates and masks them out.
  arXiv  Detail & Related papers  (2025-10-01T12:29:32Z)
• A Variational Framework for Residual-Based Adaptivity in Neural PDE Solvers and Operator Learning [3.758814046658822]
  Residual-based adaptive strategies are widely used in machine learning but remain largely.<n>We introduce a unifying variational framework that formalizes these methods by integrating convex transformations of the residual.<n>Our results provide a theoretical justification of residual-based adaptivity and establish a foundation for principled discretization and training strategies.
  arXiv  Detail & Related papers  (2025-09-17T17:34:03Z)
• Iterative Preference Learning from Human Feedback: Bridging Theory and Practice for RLHF under KL-Constraint [56.74058752955209]
  This paper studies the alignment process of generative models with Reinforcement Learning from Human Feedback (RLHF) We first identify the primary challenges of existing popular methods like offline PPO and offline DPO as lacking in strategical exploration of the environment. We propose efficient algorithms with finite-sample theoretical guarantees.
  arXiv  Detail & Related papers  (2023-12-18T18:58:42Z)
• Provable Reward-Agnostic Preference-Based Reinforcement Learning [61.39541986848391]
  Preference-based Reinforcement Learning (PbRL) is a paradigm in which an RL agent learns to optimize a task using pair-wise preference-based feedback over trajectories. We propose a theoretical reward-agnostic PbRL framework where exploratory trajectories that enable accurate learning of hidden reward functions are acquired.
  arXiv  Detail & Related papers  (2023-05-29T15:00:09Z)
• Beyond variance reduction: Understanding the true impact of baselines on policy optimization [24.09670734037029]
  We show that learning dynamics are governed by the curvature of the loss function and the noise of the gradient estimates. We present theoretical results showing that, at least for bandit problems, curvature and noise are not sufficient to explain the learning dynamics.
  arXiv  Detail & Related papers  (2020-08-31T17:52:09Z)

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.