Agents

Thought Cloning

First page

Paper summary

Imitation learning framework that learns to think as well as act.

Ask this paper

Key points

Cloning thoughts AND behavior: Clones both the actions and the internal verbal thoughts of human demonstrators, not just behavioral trajectories.

BabyAI benchmark: Demonstrated on BabyAI with substantial improvement over behavior-only cloning, especially on out-of-distribution tasks.

Interpretability bonus: Because the agent thinks in natural language, its decisions are interpretable and debuggable.

Reasoning-agent precursor: A conceptual precursor to 2024's "reasoning agents" that produce explicit thought traces before acting.

Every Monday