Tag: Reinforcement Learning
All the articles with the tag "Reinforcement Learning".
-
InfiFPO: Implicit Model Fusion via Preference Optimization in Large Language Models
Published: at 11:19 AM
85.18 🤔InfiFPO提出了一种在偏好对齐阶段进行隐式模型融合的偏好优化方法，通过序列级概率融合和优化策略，将多个源模型知识整合到枢轴模型中，显著提升了Phi-4在11个基准上的平均性能从79.95到83.33。
-
Learning Like Humans: Advancing LLM Reasoning Capabilities via Adaptive Difficulty Curriculum Learning and Expert-Guided Self-Reformulation
Published: at 11:01 AM 85.16 🤔This paper introduces Adaptive Difficulty Curriculum Learning (ADCL) and Expert-Guided Self-Reformulation (EGSR) to enhance LLM reasoning by dynamically adjusting training curricula and guiding models to reformulate expert solutions, achieving significant performance improvements over standard RL baselines on mathematical reasoning benchmarks.
-
ReMA: Learning to Meta-think for LLMs with Multi-Agent Reinforcement Learning
Published: at 11:53 AM 85.15 🤔ReMA通过多智能体强化学习分离元思考和推理过程，提升了大型语言模型在数学推理和LLM-as-a-Judge任务上的性能，尤其在分布外泛化能力上表现出色，但对超参数敏感且多轮设置存在稳定性挑战。
-
Concise Reasoning via Reinforcement Learning
Published: at 11:21 PM 85.10 🤔本文提出了一种两阶段强化学习训练策略，通过在极小数据集上分阶段优化推理能力和简洁性，显著减少大型语言模型的响应长度（最高54%），同时保持甚至提升准确性，并增强低采样强度下的鲁棒性。
-
Response-Level Rewards Are All You Need for Online Reinforcement Learning in LLMs: A Mathematical Perspective
Published: at 11:23 AM 85.08 🤔本文提出'Trajectory Policy Gradient Theorem'，从理论上证明在LLM在线强化学习中仅用响应级别奖励即可无偏估计token级奖励的策略梯度，并基于此设计了TRePO算法，简化PPO设计并具备token级建模能力。