Overview

Tree of Thoughts (ToT) generalizes Chain of Thought by allowing the model to branch out into different possibilities. It is inspired by human problem-solving techniques like trial-and-error and look-ahead search.

How it Works

  1. Thought Generation: The model generates several potential next steps.
  2. Evaluation: The model (or a separate heuristic) scores each step based on its likelihood of leading to a solution.
  3. Search: The system uses algorithms like Breadth-First Search (BFS) or Depth-First Search (DFS) to navigate the 'tree' of possibilities.

Use Cases

Extremely difficult tasks requiring strategic planning or creative exploration, such as complex math puzzles or creative writing.

Related Terms