WHAT NEW IN X11.04 WED

AI collections from X · 22 collections · 67,905 posts

Back to No.008 Gemini 4 Pro / Flash: benchmarks, hands-on tests and leaksUpdated 4 hours ago
gemini-4THREAD

Blogger starts a thread on Gemini 4 Argon safety and security evals

scaling01 announced a long thread on Gemini 4 Argon's safety evaluation, starting with GCI, Google's version of the ECI, and calling the model a possible "dirty looper."

Lisan al Gaib
@scaling01Archived 21 days agoOriginally posted 2026.10.14
THREADPosted in 2 parts

  1. im making a thread on the gemini 4 argon safety & security evaluation starting with: GCI Google's version of the ECI

    This post quotes another post: “Gemini 4 Argon exceeds human baseline on all four stealth capability evals”

    Blogger starts a thread on Gemini 4 Argon safety and security evals (thread part 1)

  2. Gemini 4 Argon might be a dirty looper

    Blogger starts a thread on Gemini 4 Argon safety and security evals (thread part 2)
Lisan al Gaib @scaling01Originally posted 2026.10.14
Open original on X
x.com/scaling01/status/2110119579829592143Archived 2026.10.14 10:05
Open original on X