Capability Calibration Through Evals
Definition
Tracking AI progress by converting tasks that models previously failed into standing test suites and rerunning them on each new model, updating expectations from the results.
Cast within
No topic-region cast yet — this appears once Capability Calibration Through Evals's compiled claims are aligned into a topic region's argument tree.
Claims mentioning Capability Calibration Through Evals (1)
Your notes
Loading notes...