风险评测 · 评测基准

LABBench2: An Improved Benchmark for AI Systems Performing Biology Research

NUWA 研究解读 · 独立复核待完成

首次发表: 2026-02-04来源核验: 2026-10-25

原始来源 ↗引用本记录 ↓化生放核 →

NUWA Research Note

研究问题研究设计与假设来源发现 NUWA 解读局限与相反证据研究与决策意义机构与角色原始资源

模型辅助整理 · 选读原始全文或官方文章中的相关章节；尚未完成逐章节审计。

研究速读

NUWA 解读：提高真实性可暴露旧格式得分上升所隐藏的缺口。

研究问题

科研能力测试能否比知识选择题更接近有用研究工作？

研究设计与假设

LABBench2含逾1,900项研究支持任务，在更真实语境比较有工具与无工具模型。

LABBench2 — Figure 2 — Figure 2 · 原文 Figure 2。请结合下方的协议范围与局限阅读。原图与来源 ↗

来源发现

作者报告相较旧基准，模型各子任务正确率差异为−26%至−46%。

工具增强效果不一致（§3）。

NUWA 解读

NUWA 解读：提高真实性可暴露旧格式得分上升所隐藏的缺口。

局限与相反证据

湿实验执行、模糊目标和跨日迭代仍被简化，验证器及领域广度有限（§4.3）。

研究与决策意义

模型建议：使用具体版本与工具条件，不据此推断操作性生物能力增益。

精确来源位置

LABBench2 An Improved Benchmark for AI Systems Performing Biology Research · AbstractSelected original resource https://arxiv.org/html/2604.09554; sha256:333fb7f67888029041cb7ff7df47ada35844806b30ea09197b5258442e65aa72; archived source capture; exact capture time not independently established
LABBench2 An Improved Benchmark for AI Systems Performing Biology Research · §3 ResultsSelected original resource https://arxiv.org/html/2604.09554; sha256:333fb7f67888029041cb7ff7df47ada35844806b30ea09197b5258442e65aa72; archived source capture; exact capture time not independently established
LABBench2 An Improved Benchmark for AI Systems Performing Biology Research · §4.3 Limitations and future workSelected original resource https://arxiv.org/html/2604.09554; sha256:333fb7f67888029041cb7ff7df47ada35844806b30ea09197b5258442e65aa72; archived source capture; exact capture time not independently established

来源核查、数据提取、编辑批准与实验复现分别记录。方法说明 →

机构与角色

研究机构 Edison Scientific FutureHouse

作者 (12)

Jon M Laurent · Albert Bou · Michael Pieler · Conor Igoe · Alex Andonian · Siddharth Narayanan · James Braza · Alexandros Sanchez Vassopoulos · Jacob L Steenwyk · Blake Lash · Andrew D White · Samuel G Rodriques

评测协议

LABBench2 →

引用与复用

引用研究发现时请引用原始出版物；引用 NUWA 的整理与解释时，单独引用本站记录及对应快照。

BibTeX ↓RIS ↓记录与来源元数据 · JSON ↓阅读覆盖 · JSON 提交更正

原始出版物 · BibTeX

@misc{labbench-two,
  title = {LABBench2: An Improved Benchmark for AI Systems Performing Biology Research},
  url = {https://arxiv.org/abs/2604.09554},
  author = {Jon M Laurent and Albert Bou and Michael Pieler and Conor Igoe and Alex Andonian and Siddharth Narayanan and James Braza and Alexandros Sanchez Vassopoulos and Jacob L Steenwyk and Blake Lash and Andrew D White and Samuel G Rodriques},
  year = {2026},
  date = {2026-02-04},
  version = {arXiv:2604.09554v2},
  note = {arXiv:2604.09554v2}
}

NUWA 记录 · BibTeX

@misc{labbench-two-nuwa,
  title = {LABBench2: An Improved Benchmark for AI Systems Performing Biology Research — NUWA research record},
  url = {https://whitzard.tech/nuwa/whitzard-index/work/labbench-two/},
  version = {2026-10-29.1},
  organization = {NUWA Frontier AI Safety Lab},
  howpublished = {Retained note edition: https://whitzard.tech/nuwa/whitzard-index/data/record-editions/2026-10-29.1--labbench-two.json},
  note = {NUWA research note. Source-reading scope and independent review status are stated on the record. Snapshot 2026-10-29.1. Retained note edition: https://whitzard.tech/nuwa/whitzard-index/data/record-editions/2026-10-29.1--labbench-two.json.}
}

证据快照 2026-10-29.1 · 索引更新 · RSS

参考文献

[1] Jon M Laurent, Albert Bou 等. LABBench2: An Improved Benchmark for AI Systems Performing Biology Research. 2026-02-04. v2 [arXiv]; Author affiliations on the original paper title page; Abstract; §3 Results