Skip navigation
New research from Databricks: LLMs Can Learn to Reason via Off-Policy RL
0:00 / 0:13

New research from Databricks: LLMs Can Learn to Reason via Off-Policy RL

Databricks AI Research

@DbrxMosaicAI55K followers

Visit on X

12,860 views

New research from Databricks: LLMs Can Learn to Reason via Off-Policy RL Optimal Advantage-based Policy Optimization with Lagged Inference policy (OAPL) shows you don’t need strict on-policy training to improve reasoning. It matches or beats Group Relative Policy Optimization (GRPO), stays stable with large policy lag, and uses ~3× fewer training generations. For Databricks customers, it’s a simpler, practical, and equally powerful approach to RL that Databricks is pioneering internally — and bringing directly to Databricks customers, so enterprises can improve agents using the same methods we use for our in-house agents, without complex infrastructure changes.

Comments

Comments from the original post on X (Twitter)

No comments yet