Goodfire's activation probes cut LLM monitoring cost by 90% on Kimi K3 with ~1% precision drop

burny_tech · x · 2026-09-19

Goodfire AI tested activation-based monitoring on Kimi K3: internal probes replace external monitors, cutting LLM monitoring cost by 90% with only 1% precision drop, making monitoring feasible at scale.

Eric Ho explains the key advantage: activation monitors run synchronously with near-zero overhead by reusing the model's own computations. Reliable reward-hacking detection could stop hacks in progress, identify broken environments, surface new failure modes, and improve training.

Related event: Goodfire: Models Know They're Reward Hacking 50-96% of the Time; Activation Monitors Catch It in Real Time(13 posts)→

Original post →

More from Safety

• Cryptographers push back: claims ECC is likely to break are not consensus — jedisct1 · 2026-11-02
• UK's 'AI sovereignty' push undermined by its own energy and copyright policy — NinaDSchick · 2026-11-02
• Auditing my MCP credential broker: the worst hole was the config file — Relevant_Mousse_9117 · 2026-11-02
• Cambricon co-founder urges shared US-China AI ethics standards for leading models — pstAsiatech · 2026-11-02
• Vitalik's "post-AI cryptography": pseudonyms are dead, per-action anonymity is in — devanshmehta · 2026-11-02
• 6 prompts to see what ChatGPT quietly knows about you — and wipe it — Aiden_Tech_Ai · 2026-11-02

Safety channel →