TROLL: Trust Regions improve Reinforcement Learning for Large Language Models

• URL: http://arxiv.org/abs/2510.03817v1
• Date: Sat, 04 Oct 2025 14:14:20 GMT
• Title: TROLL: Trust Regions improve Reinforcement Learning for Large Language Models
• Abstract summary: On-policy Reinforcement Learning (RL) with PPO-like clip objectives has become the standard choice for fine-tuning of large language models (LLMs)<n>We replace the clip objective with a novel discrete differentiable trust region projection, which provides principled token-level KL constraints.<n>Our approach, Trust Region Optimization for Large Language Models (TROLL), serves as a direct replacement for PPO-like clipping during training and does not alter the model's inference behavior.
• Score: 25.03342193102256
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: On-policy Reinforcement Learning (RL) with PPO-like clip objectives has become the standard choice for reward-based fine-tuning of large language models (LLMs). Although recent work has explored improved estimators of advantages and normalization, the clipping mechanism itself has remained untouched. Originally introduced as a proxy for principled KL-based trust regions, clipping is a crude approximation that often causes unstable updates and suboptimal performance. We replace the clip objective with a novel discrete differentiable trust region projection, which provides principled token-level KL constraints. The projection operates on a sparse subset of the model's most important token logits to balance computational cost and projection effectiveness. Our approach, Trust Region Optimization for Large Language Models (TROLL), serves as a direct replacement for PPO-like clipping during training and does not alter the model's inference behavior. Across datasets, model families, and advantage-estimation methods, TROLL consistently outperforms PPO-like clipping in terms of training speed, stability, and final success rates.

Related papers

• 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)
• Explicit Critic Guidance for Aligning Diffusion Models [27.6704888666835]
  We propose a state-aligned latent actor-critic framework for diffusion post-training.<n>Our method consistently outperforms prior group-relative RL and actor-critic baselines on single-reward and multi-reward benchmarks.
  arXiv  Detail & Related papers  (2026-05-26T22:20:51Z)
• On Predicting the Post-training Potential of Pre-trained LLMs [60.07459271263409]
  We introduce a new task of predicting post-training potential - forecasting a base model's performance before post-training.<n>We propose RuDE, a unified framework that bypasses the generation gap of base models by leveraging response discrimination.<n>Experiments demonstrate a correlation greater than 90% with post-training performance.
  arXiv  Detail & Related papers  (2026-05-12T11:33:49Z)
• Power Reinforcement Post-Training of Text-to-Image Models with Super-Linear Advantage Shaping [66.25536973294726]
  Post-training methods for text-to-image (T2I) models are prone to reward hacking.<n>Super-Linear Advantage Shaping (SLAS) reshapes the local policy space.<n>SLAS consistently surpasses the DanceGRPO baseline across multiple backbones and benchmarks.
  arXiv  Detail & Related papers  (2026-05-11T17:59:25Z)
• BandPO: Bridging Trust Regions and Ratio Clipping via Probability-Aware Bounds for LLM Reinforcement Learning [49.25750348525603]
  BandPO replaces canonical clipping with Band, a unified theoretical operator that projects trust regions into dynamic, probability-aware clipping intervals.<n>We show that BandPO consistently outperforms canonical clipping and Clip-Higher, while robustly mitigating entropy collapse.
  arXiv  Detail & Related papers  (2026-03-05T08:03:05Z)
• 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)
• BAPO: Stabilizing Off-Policy Reinforcement Learning for LLMs via Balanced Policy Optimization with Adaptive Clipping [69.74252624161652]
  We propose BAlanced Policy Optimization with Adaptive Clipping (BAPO)<n>BAPO dynamically adjusts clipping bounds to adaptively re-balance positive and negative contributions, preserve entropy, and stabilize RL optimization.<n>On AIME 2024 and AIME 2025 benchmarks, our 7B BAPO model surpasses open-source counterparts such as SkyWork-OR1-7B.
  arXiv  Detail & Related papers  (2025-10-21T12:55:04Z)
• CTR-LoRA: Curvature-Aware and Trust-Region Guided Low-Rank Adaptation for Large Language Models [23.34193229321038]
  CTR-LoRA integrates rank scheduling with stability-aware optimization.<n>Experiments on multiple open-source backbones (7B-13B) show consistent improvements over strong PEFT baselines.
  arXiv  Detail & Related papers  (2025-10-11T20:05:56Z)
• MaP: A Unified Framework for Reliable Evaluation of Pre-training Dynamics [72.00014675808228]
  Instability in Large Language Models evaluation process obscures true learning dynamics.<n>We introduce textbfMaP, a framework that integrates underlineMerging underlineand the underlinePass@k metric.<n>Experiments show that MaP yields significantly smoother performance curves, reduces inter-run variance, and ensures more consistent rankings.
  arXiv  Detail & Related papers  (2025-10-10T11:40:27Z)
• TrajBooster: Boosting Humanoid Whole-Body Manipulation via Trajectory-Centric Learning [79.59753528758361]
  We present TrajBooster, a cross-embodiment framework that leverages abundant wheeled-humanoid data to boost bipedal VLA.<n>Our key idea is to use end-effector trajectories as a morphology-agnostic interface.<n>Results show that TrajBooster allows existing wheeled-humanoid data to efficiently strengthen bipedal humanoid VLA performance.
  arXiv  Detail & Related papers  (2025-09-15T12:25:39Z)
• PITA: Preference-Guided Inference-Time Alignment for LLM Post-Training [9.093854840532062]
  PITA is a novel framework that integrates preference feedback directly into the LLM's token generation.<n> PITA learns a small preference-based guidance policy to modify token probabilities at inference time without fine-tuning.<n>We evaluate PITA across diverse tasks, including mathematical reasoning and sentiment classification.
  arXiv  Detail & Related papers  (2025-07-26T21:46:32Z)
• Lean and Mean: Decoupled Value Policy Optimization with Global Value Guidance [52.65461207786633]
  Policy-based Reinforcement Learning from Human Feedback is essential for aligning large language models with human preferences.<n>It requires joint training of an actor and critic with a pretrained, fixed reward model for guidance.<n>We propose textbfDecoupled Value Policy Optimization (DVPO), a lean framework that replaces traditional reward modeling with a pretrained emphglobal value model (GVM)
  arXiv  Detail & Related papers  (2025-02-24T08:11:33Z)
• SALSA: Soup-based Alignment Learning for Stronger Adaptation in RLHF [22.88031166401938]
  This paper presents SALSA, a novel approach designed to overcome limitations by creating a more flexible and better located reference model. We show that SALSA fosters better exploration, achieving higher rewards and improving model robustness, out-of-distribution, and performance.
  arXiv  Detail & Related papers  (2024-11-04T04:53:43Z)
• Aligning Large Language Models via Fine-grained Supervision [20.35000061196631]
  Pre-trained large-scale language models (LLMs) excel at producing coherent articles, yet their outputs may be untruthful, toxic, or fail to align with user expectations. Current approaches focus on using reinforcement learning with human feedback to improve model alignment. We propose a method to enhance LLM alignment through fine-grained token-level supervision.
  arXiv  Detail & Related papers  (2024-06-04T20:21:45Z)
• Towards Stable Machine Learning Model Retraining via Slowly Varying Sequences [6.067007470552307]
  We propose a model-agnostic framework for finding sequences of models that are stable across retraining iterations.<n>We develop a mixed-integer optimization formulation that is guaranteed to recover optimal models.<n>We find that, on average, a 2% reduction in predictive power leads to a 30% improvement in stability.
  arXiv  Detail & Related papers  (2024-03-28T22:45:38Z)
• Learn from the Past: A Proxy Guided Adversarial Defense Framework with Self Distillation Regularization [53.04697800214848]
  Adversarial Training (AT) is pivotal in fortifying the robustness of deep learning models. AT methods, relying on direct iterative updates for target model's defense, frequently encounter obstacles such as unstable training and catastrophic overfitting. We present a general proxy guided defense framework, LAST' (bf Learn from the Pbf ast)
  arXiv  Detail & Related papers  (2023-10-19T13:13:41Z)
• Distributionally Robust Post-hoc Classifiers under Prior Shifts [31.237674771958165]
  We investigate the problem of training models that are robust to shifts caused by changes in the distribution of class-priors or group-priors. We present an extremely lightweight post-hoc approach that performs scaling adjustments to predictions from a pre-trained model.
  arXiv  Detail & Related papers  (2023-09-16T00:54:57Z)
• Fine-Tuning Language Models with Advantage-Induced Policy Alignment [80.96507425217472]
  We propose a novel algorithm for aligning large language models to human preferences. We show that it consistently outperforms PPO in language tasks by a large margin. We also provide a theoretical justification supporting the design of our loss function.
  arXiv  Detail & Related papers  (2023-06-04T01:59:40Z)

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.