8 papers
The Mirage of Optimizing Training Policies: Monotonic Inference Policies as the Real Objective for LLM Reinforcement Learning
Jing Liang, Hongyao Tang, Yi Ma +9
Reinforcement learning (RL) has gained growing attention in large language model (LLM) post-training, yet RL training remains fragile and can suffer from instability or collapse. O…
From Seeing to Doing: Bridging Reasoning and Decision for Robotic Manipulation
Yifu Yuan, Haiqin Cui, Yibin Chen +7
Achieving generalization in robotic manipulation remains a critical challenge, particularly for unseen scenarios and novel tasks. Current Vision-Language-Action (VLA) models, while…
Key Decision-Makers in Multi-Agent Debates: Who Holds the Power?
Qian Zhang, Yan Zheng, Jinyi Liu +2
Recent studies on LLM agent scaling have highlighted the potential of Multi-Agent Debate (MAD) to enhance reasoning abilities. However, the critical aspect of role allocation strat…
DualRAG: A Dual-Process Approach to Integrate Reasoning and Retrieval for Multi-Hop Question Answering
Rong Cheng, Jinyi Liu, Yan Zheng +6
Multi-Hop Question Answering (MHQA) tasks permeate real-world applications, posing challenges in orchestrating multi-step reasoning across diverse knowledge domains. While existing…
Squeeze the Soaked Sponge: Efficient Off-policy Reinforcement Finetuning for Large Language Model
Jing Liang, Hongyao Tang, Yi Ma +5
Reinforcement Learning (RL) has demonstrated its potential to improve the reasoning ability of Large Language Models (LLMs). One major limitation of most existing Reinforcement Fin…
Prioritized Trajectory Replay: A Replay Memory for Data-driven Reinforcement Learning
Jinyi Liu, Yi Ma, Jianye Hao +4
In recent years, data-driven reinforcement learning (RL), also known as offline RL, have gained significant attention. However, the role of data sampling techniques in offline RL h…