Sham Kakade
@ShamKakade6 • 19,739 subscribers
Harvard Professor. Full stack ML and AI. Co-director of the Kempner Institute for the Study of Artificial and Natural Intelligence.
Videos
1/ Au revoir, RLVR. New work: EBFT (Energy-Based Fine-Tuning), a post-training method that directly optimizes the long-horizon behavior of model generations, addressing SFT’s deployment-time error amplification without relying on sparse, task-specific rewards.
Sham Kakade267,650 Aufrufe • vor 6 Monaten
Keine weiteren Inhalte verfügbar