LiTS: A Modular Framework for LLM Tree Search

TL;DR

LiTS is a modular Python framework that enables flexible LLM reasoning via tree search, supporting domain-specific components and algorithms, and revealing insights about policy diversity in infinite action spaces.

Contribution

Introduces a reusable, modular Python framework for LLM reasoning with tree search, allowing easy extension and demonstrating component and algorithm orthogonality.

Findings

Components are reusable across algorithms within each task type.

Algorithms work across all components and domains.

Policy diversity limits effectiveness in infinite action spaces.

Abstract

LiTS is a modular Python framework for LLM reasoning via tree search. It decomposes tree search into three reusable components (Policy, Transition, and RewardModel) that plug into algorithms like MCTS and BFS. A decorator-based registry enables domain experts to extend to new domains by registering components, and algorithmic researchers to implement custom search algorithms. We demonstrate composability on MATH500 (language reasoning), Crosswords (environment planning), and MapEval (tool use), showing that components and algorithms are orthogonal: components are reusable across algorithms within each task type, and algorithms work across all components and domains. We also report a mode-collapse finding: in infinite action spaces, LLM policy diversity (not reward quality) is the bottleneck for effective tree search. A demonstration video is available at https://youtu.be/nRGX43YrR3I. The package is released under the Apache 2.0 license at https://github.com/xinzhel/lits-llm, including installation instructions and runnable examples that enable users to reproduce the demonstrated workflows.