Research
Research & blog
Privileged Self-Distillation
Post-training method that builds privileged repair hints from failed rollouts and distills back into the policy.
July 2026ReadMeta-Reward: Reward Modeling as Harness Optimization
Optimizing LLM-judge evaluator harnesses to build better agent reward models.
May 2026ReadMeta-Agent: Continual Learning for Agents
An open-source library that automatically improves agent harnesses from production traces.
May 2026ReadCanvas Labs
How deploying Canvas led us to build the continual-improvement layer for production AI agents.
April 2026Read