Event ·Official first-hand

研究显示前沿大模型在架构设计上存在相似性

这篇论文测试了OpenAI、Anthropic、xAI和Google DeepMind的六种前沿模型，发现它们在架构设计上存在相似性。当用‘学校听众’的框架提示时，这些模型反复收敛到一种共享的架构模式，包括持久化隐状态、自适应计算、内存、专家路由、验证、停止控制和延迟解码。GPT-5.6 Sol和GPT-6 Astra的架构设计存在重叠，这表明它们可能独立地想象着相似的架构未来。

Current read

这篇论文很有意思，它测试了不同公司的前沿大模型，发现它们在设计架构时，会倾向于一种相似的思路，比如都考虑了持久化隐状态和专家路由这些要素。

11 sources44° AI heatLast updated 2026/9/13 21:32:35

Evidence

11 sources

Primary sourcearXiv: Google DeepMind

2026/9/13 21:32:35View source

Related source 1The Rundown AI

2026/9/15 16:32:06View source

Related source 2theverge

2026/9/14 22:59:41View source

Related source 3techcrunch

2026/9/15 15:47:03View source

Related source 4AI Engineer

2026/9/15 17:10:00View source

Related source 5@OpenAIDevs

2026/9/12 08:09:11View source

Related source 6Decoder

2026/9/12 13:10:11View source

Related source 7宝玉

2026/9/12 18:10:51View source

Related source 8IT之家

2026/9/13 09:20:17View source

Related source 9小互

2026/9/14 00:32:03View source

Related source 10Gary Marcus

2026/9/14 13:09:08View 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