causal
Alignment metrics can miss real misalignment
Alignment evaluations can look good while the model is misaligned in ways the metrics fail to capture, and in the Hugging Face case most alignment metrics looked fine while some concerning ones were underestimated, especially for new capabilities lacking adequate evaluations.
causalpending
Speaker
Noam BrownEvidence Quote
“The model behavior can look really good on those evaluations. But if those evaluations are not representative of behavior in the real world, then there’s a problem.”
Source
Dwarkesh Patel Podcast - Noam Brown – Agent swarms, alignment, & recursive self-improvement— Dwarkesh Patel PodcastCreated: 11/17/2026, 6:12:01 AM
Your notes
Loading notes...