7 papers
Reward Modeling for Reinforcement Learning-Based LLM Reasoning: Design, Challenges, and Evaluation
Pei-Chi Pan, Yingbin Liang, Sen Lin
Large Language Models (LLMs) demonstrate transformative potential, yet their reasoning remains inconsistent and unreliable. Reinforcement learning (RL)-based fine-tuning is a key m…
Constraint-Rectified Training for Efficient Chain-of-Thought
Qinhang Wu, Sen Lin, Ming Zhang +2
Chain-of-Thought (CoT) has significantly enhanced the reasoning capabilities of Large Language Models (LLMs), especially when combined with reinforcement learning (RL) based post-t…
A Comprehensive Survey of Mixture-of-Experts: Algorithms, Theory, and Applications
Siyuan Mu, Sen Lin
Artificial intelligence (AI) has achieved astonishing successes in many domains, especially with the recent breakthroughs in the development of foundational large models. These lar…
Towards Fast Safe Online Reinforcement Learning via Policy Finetuning
Keru Chen, Honghao Wei, Zhigang Deng +1
The high costs and risks involved in extensive environment interactions hinder the practical application of current online safe reinforcement learning (RL) methods. While offline s…
Mixture-of-Transformers Learn Faster: A Theoretical Study on Classification Problems
Hongbo Li, Qinhang Wu, Sen Lin +2
Mixture-of-Experts (MoE) models improve transformer efficiency but lack a unified theoretical explanation, especially when both feed-forward and attention layers are allowed to spe…
Unlocking the Power of Rehearsal in Continual Learning: A Theoretical Perspective
Junze Deng, Qinhang Wu, Peizhong Ju +3
Rehearsal-based methods have shown superior performance in addressing catastrophic forgetting in continual learning (CL) by storing and training on a subset of past data alongside…