← 研踪 PaperTrailarXiv:2608.20317 · v2
CMASS: A ClimbMix Agentic Search Suite for Evaluating Search Agents Across Heterogeneous Tasks
首次提交 2026-08-21 01:54
版本更新 2026-10-27 00:55
雷达观测 2026-10-29 00:35
摘要
Evaluating advances in agentic search requires testing whether methods generalize across tasks and corpus scale. Changing web search services impede controlled comparisons, while small corpora assembled around benchmark questions provide a limited test of generalization. We introduce CMASS, the ClimbMix Agentic Search Suite, which projects three search benchmarks---BrowseComp-Plus, Total Recall QA, and DeepSearchQA---and the multi-hop question-answering dataset HotpotQA onto a shared, fixed corpus. The target, ClimbMix, is a pre-training corpus of 553M documents and approximately 400B tokens, assembled independently of the benchmark questions. Our projection pipeline verifies grounding, answer uniqueness, and set completeness through independent agent audits, controlled revision, and human review. The 423 questions span entity identification, multi-hop reasoning, exhaustive answer sets, and aggregation. Evaluations of three models expose substantial headroom for open-weight agents and show that high answer accuracy can coexist with incomplete evidence coverage. By enabling measurement of accuracy alongside evidence coverage, search effort, and closed-book performance, CMASS supports developing agents that find sufficient evidence and produce complete answers efficiently. The pipeline code and dataset are available at https://cmass.tech
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v2 · 2026-10-27 00:55 · 最新观测版本
CMASS: A ClimbMix Agentic Search Suite for Evaluating Search Agents Across Heterogeneous Tasks
Evaluating advances in agentic search requires testing whether methods generalize across tasks and corpus scale. Changing web search services impede controlled comparisons, while small corpora assembled around benchmark questions provide a limited test of generalization. We introduce CMASS, the ClimbMix Agentic Search Suite, which projects three search benchmarks---BrowseComp-Plus, Total Recall QA, and DeepSearchQA---and the multi-hop question-answering dataset HotpotQA onto a shared, fixed corpus. The target, ClimbMix, is a pre-training corpus of 553M documents and approximately 400B tokens, assembled independently of the benchmark questions. Our projection pipeline verifies grounding, answer uniqueness, and set completeness through independent agent audits, controlled revision, and human review. The 423 questions span entity identification, multi-hop reasoning, exhaustive answer sets, and aggregation. Evaluations of three models expose substantial headroom for open-weight agents and show that high answer accuracy can coexist with incomplete evidence coverage. By enabling measurement of accuracy alongside evidence coverage, search effort, and closed-book performance, CMASS supports developing agents that find sufficient evidence and produce complete answers efficiently. The pipeline code and dataset are available at https://cmass.tech
查看此版本解读 →查看此版本原文 ↗