Picture for Sándor Toth

Sándor Toth

Tapered Off-Policy REINFORCE: Stable and efficient reinforcement learning for LLMs

Add code
Mar 19, 2025
Viaarxiv icon