RL Scaling Laws w/ Devvrit Khatri
NeurIPS
OVERVIEW
Join us for our first reading group event in San Diego at NeurIPS with Devvrit Khatri (UT Austin, Meta Superintelligence). Lunch will be catered for all attendees!
For this session, we'll be discussing Devvrit's breakthrough paper in the RL Scaling Laws domain:
The Art of Scaling Reinforcement Learning Compute for LLMs (link)
We'll gather ~25-30 folks of different research backgrounds together to go through the key sections of the paper, discuss results, and understand important research outcomes.