English
Related papers

Related papers: A Step Back: Prefix Importance Ratio Stabilizes Po…

200 papers

We propose DiFFPO, Diffusion Fast and Furious Policy Optimization, a unified framework for training masked diffusion large language models (dLLMs) to reason not only better (furious), but also faster via reinforcement learning (RL). We…

Machine Learning · Computer Science 2026-01-13 Hanyang Zhao , Dawen Liang , Wenpin Tang , David Yao , Nathan Kallus

Reinforcement learning (RL) has emerged as a promising strategy for improving the reasoning capabilities of language models (LMs) in domains such as mathematics and coding. However, most modern RL algorithms were designed to target robotics…

Artificial Intelligence · Computer Science 2025-05-26 Lianghuan Huang , Shuo Li , Sagnik Anupam , Insup Lee , Osbert Bastani

Diffusion large language models (dLLMs) are promising alternatives to autoregressive large language models (AR-LLMs), as they potentially allow higher inference throughput. Reinforcement learning (RL) is a crucial component for dLLMs to…

Machine Learning · Computer Science 2026-02-24 Yuchen Zhu , Wei Guo , Jaemoo Choi , Petr Molodyk , Bo Yuan , Molei Tao , Yongxin Chen

Alignment of large language models (LLMs) typically involves training a reward model on preference data, followed by policy optimization with respect to the reward model. However, optimizing policies with respect to a single reward model…

Machine Learning · Computer Science 2025-07-23 Debangshu Banerjee , Kintan Saha , Aditya Gopalan

Large language models (LLMs) exhibit strong reasoning capabilities but typically require expensive post-training to reach high performance. Recent test-time alignment methods offer a lightweight alternative, but have been explored mainly…

Computation and Language · Computer Science 2026-03-20 Arushi Rai , Qiang Zhang , Hanqing Zeng , Yunkai Zhang , Dipesh Tamboli , Xiangjun Fan , Zhuokai Zhao , Lizhu Zhang

Training Large Language Models (LLMs) to reason often relies on Reinforcement Learning (RL) with task-specific verifiers. However, many real-world reasoning-intensive tasks lack verifiers, despite offering abundant expert demonstrations…

Machine Learning · Computer Science 2025-12-10 Locke Cai , Ivan Provilkov

Reinforcement learning for multi-step reasoning with large language models (LLMs) typically relies on sparse terminal rewards, which creates a poorly conditioned credit-assignment problem: the final feedback is propagated uniformly across…

Machine Learning · Computer Science 2026-05-26 Fei Ding , Yongkang Zhang , youwei wang , Zijian Zeng

Pre-trained Language Models (PLMs) can be accurately fine-tuned for downstream text processing tasks. Recently, researchers have introduced several parameter-efficient fine-tuning methods that optimize input prompts or adjust a small number…

Computation and Language · Computer Science 2024-06-07 Saeed Najafi , Alona Fyshe

Reinforcement Learning (RL) has emerged as the key driver for post-training complex reasoning in Large Language Models (LLMs), yet online RL introduces significant instability and computational overhead. Offline RL offers a compelling…

Computation and Language · Computer Science 2026-04-06 Minjae Oh , Yunho Choi , Dongmin Choi , Yohan Jo

Traditional on-policy Reinforcement Learning with Verifiable Rewards (RLVR) frameworks suffer from experience waste and reward homogeneity, which directly hinders learning efficiency on difficult samples during large language models…

Artificial Intelligence · Computer Science 2026-03-17 Xu Wan , Yansheng Wang , Wenqi Huang , Mingyang Sun

Reinforcement learning (RL) plays a crucial role in shaping the behavior of large language and reasoning models (LLMs/LRMs). However, it often produces brittle and unstable policies, leading to critical failures such as spurious reasoning,…

Artificial Intelligence · Computer Science 2025-07-29 Xingcheng Xu

We study the problem of training and fine-tuning expressive policies with online reinforcement learning (RL) given an offline dataset. Training expressive policy classes with online RL present a unique challenge of stable value…

Machine Learning · Computer Science 2026-05-01 Perry Dong , Qiyang Li , Dorsa Sadigh , Chelsea Finn

Large language models (LLMs) have recently advanced in reasoning when optimized with reinforcement learning (RL) under verifiable rewards. Existing methods primarily rely on outcome-based supervision to strengthen internal LLM reasoning,…

Artificial Intelligence · Computer Science 2026-05-29 Siyao Song , Cong Ma , Zhihao Cheng , Shiye Lei , Minghao Li , Ying Zeng , Huaixiao Tou , Kai Jia

The recent remarkable progress of deep reinforcement learning (DRL) stands on regularization of policy for stable and efficient learning. A popular method, named proximal policy optimization (PPO), has been introduced for this purpose. PPO…

Machine Learning · Computer Science 2023-07-04 Taisuke Kobayashi

Large language models (LLMs) are increasingly used as rerankers in information retrieval, yet their ranking behavior can be steered by small, natural-sounding prompts. To expose this vulnerability, we present Rank Anything First (RAF), a…

Computation and Language · Computer Science 2025-10-09 Tiancheng Xing , Jerry Li , Yixuan Du , Xiyang Hu

Offline reinforcement learning (RL) enables policy learning from pre-collected datasets, avoiding costly and risky online interactions, but it often struggles with long-horizon tasks involving sparse rewards. Existing goal-conditioned and…

Machine Learning · Computer Science 2026-01-14 Chengyang Gu , Yuxin Pan , Hui Xiong , Yize Chen

RL-based post-training of language models is almost exclusively done using on-policy methods such as PPO. These methods cannot learn from arbitrary sequences such as those produced earlier in training, in earlier runs, by human experts or…

Machine Learning · Computer Science 2025-03-10 Taco Cohen , David W. Zhang , Kunhao Zheng , Yunhao Tang , Remi Munos , Gabriel Synnaeve

Theoretical guarantees in reinforcement learning (RL) are known to suffer multiplicative blow-up factors with respect to the misspecification error of function approximation. Yet, the nature of such \emph{approximation factors} --…

Machine Learning · Computer Science 2023-12-18 Philip Amortila , Nan Jiang , Csaba Szepesvári

Reinforcement Learning with Verifiable Rewards (RLVR) is an essential paradigm that enhances the reasoning capabilities of Large Language Models (LLMs). However, existing methods typically rely on static policy optimization schemes that…

Computation and Language · Computer Science 2026-05-08 Yiming Huang , Zhenbo Shi , Shuzheng Gao , Cuiyun Gao , Peiyi Han , Chuanyi Liu

Preference optimization (PO) is indispensable for large language models (LLMs), with methods such as direct preference optimization (DPO) and proximal policy optimization (PPO) achieving great success. A common belief is that DPO is…

Machine Learning · Computer Science 2026-05-18 Yue Wang , Qizhou Wang , Zizhuo Zhang , Gang Niu , Bo Han , Masashi Sugiyama