Search-o1: Agentic Search-Enhanced Large Reasoning Models

9 Jan 2025  ยท  Xiaoxi Li, Guanting Dong, Jiajie Jin, Yuyao Zhang, Yujia Zhou, Yutao Zhu, Peitian Zhang, Zhicheng Dou ยท

Large reasoning models (LRMs) like OpenAI-o1 have demonstrated impressive long stepwise reasoning capabilities through large-scale reinforcement learning. However, their extended reasoning processes often suffer from knowledge insufficiency, leading to frequent uncertainties and potential errors. To address this limitation, we introduce \textbf{Search-o1}, a framework that enhances LRMs with an agentic retrieval-augmented generation (RAG) mechanism and a Reason-in-Documents module for refining retrieved documents. Search-o1 integrates an agentic search workflow into the reasoning process, enabling dynamic retrieval of external knowledge when LRMs encounter uncertain knowledge points. Additionally, due to the verbose nature of retrieved documents, we design a separate Reason-in-Documents module to deeply analyze the retrieved information before injecting it into the reasoning chain, minimizing noise and preserving coherent reasoning flow. Extensive experiments on complex reasoning tasks in science, mathematics, and coding, as well as six open-domain QA benchmarks, demonstrate the strong performance of Search-o1. This approach enhances the trustworthiness and applicability of LRMs in complex reasoning tasks, paving the way for more reliable and versatile intelligent systems. The code is available at \url{https://github.com/sunnynexus/Search-o1}.

PDF Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Uses Extra
Training Data
Benchmark
Mathematical Reasoning AIME24 Search-o1 Acc 56.7 # 4
Mathematical Reasoning AMC23 Search-o1 Acc 85 # 1
GPQA Search-o1 Accuracy 63.6 # 3
Code Generation Livecodebench Search-o1 Acc 33 # 2
Mathematical Reasoning MATH500 Search-o1 Acc 86.4 # 1
Question Answering Natural Questions Search-o1 EM 34 # 30
Question Answering TriviaQA Search-o1 F1 74.1 # 6

Methods


No methods listed for this paper. Add relevant methods here