BrowseComp

'BrowseComp' is a benchmark dataset that contains synthesized samples used to evaluate the performance of search agents, particularly in their ability to perform complex, multi-hop reasoning tasks.

Papers using BrowseComp (7)

SearchSwarm: Towards Delegation Intelligence in Agentic LLMs for Long-Horizon Deep Research2026OpenSeeker: Democratizing Frontier Search Agents by Fully Open-Sourcing Training Data2026OpenSeeker-v2: Pushing the Limits of Search Agents with Informative and High-Difficulty Trajectories2026BrowseComp-Plus: A More Fair and Transparent Evaluation Benchmark of Deep-Research Agent2025SSRL: Self-Search Reinforcement Learning2025A^2FM: An Adaptive Agent Foundation Model for Tool-Aware Hybrid Reasoning2025WebLeaper: Empowering Efficiency and Efficacy in WebAgent via Enabling Info-Rich Seeking2025