Nicolas Le Roux

Tapered Off-Policy REINFORCE - Stable and efficient reinforcement learning for large language models

VinePPO: Refining Credit Assignment in RL Training of LLMs

VinePPO: Unlocking RL Potential For LLM Reasoning Through Refined Credit Assignment

fPLSA: Learning Semantic Structures in Document Collections Using Foundation Models

Improving Context-Aware Preference Modeling for Language Models