WHAT NEW IN X11.07 SAT

X 上的 AI 合集档案 · 23 个合集 · 70,791 条

gemini-4POST

24 小时实测:Sol 6.1 查错 7/9 轮,Gemini 4 Argon 仅 2/9

作者用 24 小时复杂任务对比 Sol 6.1 与 Gemini 4 Argon,称 Argon 只提出 2 轮修正且其中一次是幻觉。

Da7em
@Da7_Tech21 天前收录原推发布于 2026.10.16
Da7em @Da7_Tech原推发布于 2026.10.16
在 X 上打开原推

I tested Sol 6.1 against Gemini 4 Argon over 24 hours of massive, complex work. Sol caught errors in 7 out of 9 rounds with 100% accuracy. Argon only requested fixes in 2 out of 9 rounds. One was good, and the other was pure hallucination.

x.com/da7_tech/status/2110787273109213439采集于 2026.10.16 10:09