Primary sourceDAIR.AI
View sourceEvent ·Multi-source
Anthropic 提出 Agent4Sci:代码智能体在 20 项科学 ML 基准上超越已发表最优模型
Anthropic 发布 Agent4Sci 论文,让前沿编程智能体在 11 个领域的 20 项科学 ML 基准上挑战已发表最优模型,结果全部超越。Claude Opus 平均改进 10.4%,且没有任何一次运行被标记为学术不端。分析显示收益几乎全部来自标准 ML 工程技巧,智能体常误判新颖性、闲置算力并陷入小范围参数扫描。加入科研构思、实验管理、定时清单三个轻量技能后,Claude Opus 的提升从 10.4% 升至 20.1%,Codex 从 4.4% 升至 8.1%。模型本身不变,研究脚手架让收益大约翻倍。
Current read
Anthropic 让 Claude Opus 和 Codex 去刷 20 个科学 ML 基准,全部打败已发表最优结果,加三个技能后提升直接翻倍,做 AI 科研系统的必看。
11 sources68° AI heatLast updated 2026/10/22 14:24:02
Evidence
11 sourcesRelated source 1rohanpaul_ai
View sourceRelated source 2swyx (AI Engineer)
View sourceRelated source 3IT之家
View sourceRelated source 4TestingCatalog
View sourceRelated source 5kimmonismus
View sourceRelated source 6theverge
View sourceRelated source 7AIGCLINK
View sourceRelated source 8Gorden Sun
View sourceRelated source 9岚叔
View sourceRelated source 10Decoder
View sourceConflict check
Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.