WHAT NEW IN X10.28 WED

AI collections from X · 22 collections · 58,924 posts

gemini-4VIDEO

Gemini 4 Argon rerun with Antigravity Harness jumps to second

After ~20% of test runs ended early on fallback, rerunning with Antigravity Harness put Opus 5.5 at 58%, Gemini 4 Argon at 54% and GPT-6 Astra at 44%, lifting Argon to second place.

Patrick 老派
@laopai1990Archived 1 day agoOriginally posted 2026.10.26
Patrick 老派 @laopai1990Originally posted 2026.10.26
Open original on X

Gemini 4 Argon 这次真被冤枉了?? 原来此前约 20% 的测试运行遇到 fallback 就直接结束, 换成 Antigravity Harness 重新测试: Opus 5.5:58% Gemini 4 Argon:54% GPT-6 Astra:44% Argon 直接冲到第二! 所以我们之前到底是在测模型能力,还是在测谁的 Harness 更靠谱??这情况发生好几次了。。。 https://x.com/wenhaocha1/status/2114559544156205413/video/1

x.com/laopai1990/status/2114661902563590213Archived 2026.10.27 09:54