Eric Ho
@eric_ho
co-founder / ceo @GoodfireAI - AI interpretability research company
Joined September 2011
581 Following 5.2K Followers
activation monitoring is faster, cheaper, and more effective than just running external monitors
one advantage is that activation monitors can run synchronously for almost no additional overhead because they can reuse the computations of the model
Show more
On Kimi K3, our probes cut the cost of LLM monitoring by 90% with a ~1% drop in precision - making monitoring feasible at scale.
Reliably detecting reward hacking could let us stop hacks in progress, identify broken environments, discover new failure modes, and fix training.
Show more
