发表机构
Bogazici University; Codeway Studios; Friedrich-Alexander-Universität Erlangen-Nürnberg; Koç University; KUIS AI Lab（博阿齐奇大学; Codeway工作室; 埃尔朗根-纽伦堡大学; 科克大学; KUIS人工智能实验室）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
介绍开源Python库TreeThink，用于神经定理证明的模块化全异步树搜索，集成多种方法技术，支持多种语言，连接REPL服务器，经评估在miniF2F和MATH500上有跨语言证明搜索等优势及异步加速。
AI 中文摘要
树搜索算法有助于在神经定理证明中系统地探索证明空间。现有的大语言模型树搜索库主要针对自然语言推理，未与形式验证器原生集成，而定理证明系统常依赖特定任务搜索实现。我们引入了TreeThink，一个用于神经定理证明中模块化、全异步树搜索的开源Python库。它将既定树搜索方法与基于vLLM的推理管道及多种节点评估技术集成，支持Lean~4、Rocq和Isabelle/HOL以及自然语言，直接连接到每种语言的读-求值-输出循环（REPL）服务器进行实时验证和证明状态提取。我们在miniF2F和MATH500上评估了TreeThink，展示了跨语言形式证明搜索、自然语言推理支持以及异步执行带来的高达6.3倍的挂钟加速。源代码根据MIT许可在这个https URL发布，该库可作为可下载包在这个https URL获取。
英文摘要
Tree search algorithms enable systematic exploration of the proof space in neural theorem proving. Existing LLM tree search libraries primarily target natural language reasoning and do not provide native integration with formal verifiers, while theorem proving systems often rely on task-specific search implementations. We introduce TreeThink, an open-source Python library for modular, fully asynchronous tree search in neural theorem proving. It integrates established tree search methods with vLLM-based inference pipelines and diverse node evaluation techniques, ranging from lightweight heuristics to neural evaluators. We support Lean~4, Rocq, and Isabelle/HOL alongside natural language. It connects directly to each language's Read-Eval-Print Loop (REPL) server for real-time verification and proof state extraction. We evaluate TreeThink on miniF2F and MATH500, demonstrating cross-language formal proof search, natural language reasoning support, and up to 8.0$\times$ wall-clock speedup from asynchronous execution. Source code is released under the MIT license at https://github.com/GGLAB-KU/treethink , and the library is accessible as a downloadable package at https://pypi.org/project/treethink/ .
CommentsEMNLP 2026 System Demonstrations
