The evidence behind the guides

Sources

Find a study, check what it supports, and follow it back to a concern.

TopicCollection

Sources for Evaluation cost & inaccessibility

2 sources

Guide citations are still under editorial review.

Longpre and colleagues · 2024 Research proposal

A Safe Harbor for AI Evaluation and Red Teaming (original source, new tab)

Proposes legal and technical protections for independent safety research; the proposal is not itself binding policy.

Kapoor and colleagues · 2024 Research analysis

AI Agents That Matter (original source, new tab)

Examines agent evaluation, cost, reproducibility and benchmark overfitting; concerns the benchmarks studied.

Sources retain their original publishers’ rights. Inclusion does not imply endorsement. Saved sources stay in this browser on this device. Dates and source descriptions come from the existing editorial records.