Gemini 4 Argon ships Google's most thorough safety report yet, closing the word-count gap with Anthropic

NeelNanda5 · x · 2026-10-14

Google DeepMind's Neel Nanda says the Gemini 4 Argon safety report is Google's most thorough to date, praising its loss-of-control, alignment and monitorability assessments. A chart of system card + safety report lengths shows Argon at 62k words — matching Anthropic's latest and far above OpenAI (27k) and xAI (5k).

Related event: Google publishes Gemini 4 Argon safety report; model shown to detect its own evaluations(10 posts)→

Original post →

More from Models

• Snap launches as first fully sub-quadratic LLM in production with constant compute per token — ryanallred · 2026-10-14
• xAI says legacy SuperGrok Heavy plans can't be restored after billing merge mistake — RexDouglass · 2026-10-14
• HyperThink uses a hypernetwork to replace long chain-of-thought with one parameter update — Donggyun Kim · 2026-10-14
• Uberti: background voice cancellation is underparameterized for noisy audio — juberti · 2026-10-14
• Gemini 4 Argon leaps in 3D reconstruction: 81.2 on BuildingBench at ~1/5 cost — Lianhuiq · 2026-10-14
• WebRTC author Justin Uberti: diarization is the wrong path, train LLMs on noise — juberti · 2026-10-14

Models channel →