Concept detail
unreliable capability evaluation
0 tagged videos · 1 connection videos · degree 0
Publication timeline
- 2026-071 video
Source videos
Observed neighbors
human preference metrics
biases — optimization for preference makes satisfaction and preference scores overstate real correctness
1 source videos
Jev CEO: I made ChatGPT, now I'm building what's nextRelated tagged insights
No core insights are tagged with this concept.