EvoBrowseComp: Benchmarking Search Agents on Evolving Knowledge
For researchers evaluating large language models with search tools, EvoBrowseComp addresses the problem of test-set contamination and parametric memorization in static benchmarks, providing a continuously updated evaluation that prevents data leakage.
EvoBrowseComp introduces an evolving benchmark of 400 English and 400 Chinese contamination-free complex questions for search agents, using a three-agent collaborative framework to synthesize questions from live-web knowledge. The benchmark demonstrates high difficulty requiring broad horizontal search, establishing a scalable paradigm for auto-updatable evaluation.
Search Agents -- large language models augmented with search tools -- have intensified the need for future-proof evaluation benchmarks. Existing benchmarks such as BrowseComp rely on static knowledge, making them vulnerable to test-set contamination and parametric memorization. Consequently, models can achieve high scores through fact recall rather than genuine retrieval, obscuring true browsing competence via reasoning shortcuts. In this paper, we introduce EvoBrowseComp, an evolving benchmark of 400 English and 400 Chinese contamination-free complex questions synthesized via live-web traversal. To collect these questions, we design a three-agent collaborative framework: (1) a QA synthesis agent that retrieves fresh knowledge from the live web to synthesize QA pairs; (2) an information filtering agent that filters retrieved knowledge in terms of credibility and popularity to block parametric shortcuts; and (3) a high-level guidance agent that formalizes questions into reasoning graphs to reduce logical redundancy and shortcuts in synthesized QA pairs. Because the framework supports fully automated synthesis, EvoBrowseComp can be regularly updated to prevent data contamination and maintain temporal freshness. Extensive experiments confirm its great difficulty, requiring broad horizontal search. It establishes a scalable paradigm for auto-updatable, high-difficulty benchmarking that keeps pace with both evolving world knowledge and advancing agent capabilities.