Event ·Multi-source

Anthropic发布新研究，Claude模型能力再提升

Anthropic近日发布了一项新研究，聚焦于提升Claude模型在复杂任务中的表现。该研究引入了新的训练方法，使Claude在推理和代码生成任务上取得了显著进步。在多个基准测试中，Claude的得分超越了前代版本，接近GPT-4的水平。研究团队还公开了部分技术细节，为后续模型优化提供了方向。

Current read

Anthropic发了新研究，Claude在推理和代码上又强了，接近GPT-4，想了解细节的可以看看。

11 sources12° AI heatLast updated 2026/8/12 10:30:29

Evidence

11 sources

Primary sourceThe Rundown AI

2026/8/12 10:30:29View source

Related source 1arXiv cs.AI

2026/8/10 17:24:50View source

Related source 2Gary Marcus

2026/8/10 18:26:07View source

Related source 3Anthropic: Research

2026/8/10 20:31:42View source

Related source 4量子位

2026/8/11 05:12:20View source

Related source 5IT之家

2026/8/11 10:35:17View source

Related source 6Yangyi

2026/8/12 03:19:22View source

Related source 7小互

2026/8/13 14:15:19View source

Related source 8techcrunch

2026/8/13 18:28:14View source

Related source 9宝玉

2026/8/10 19:48:03View source

Related source 10Paul Graham

2026/8/10 21:16:30View source

Conflict 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.

Open the story