Tree of Thoughts: Deliberate Problem Solving with Large Language Models

Generalised chain-of-thought to tree-search exploration over units of thought using breadth-first and depth-first search with lookahead and backtracking.

Directly guides architectural decisions, alignment strategy, and serving infrastructure for task-autonomous-agents, task-text-generation.

Assumptions

• Empirical distribution regularity holds and target domain adheres to pretraining linguistic/visual support

Limitations

• Resource scaling, inference memory requirements, and alignment robustness vary with model size and hardware topology