inspect-ai
UK AISI's framework for safety evaluations of large language models, with task-based scaffolding and solver pipelines.
- Score
- 98.7432 signals
- Evidence
- 2,683 stars · 685 forks
- Last commit
- as last read from GitHub; most reads are from 2 Sep 2026 or later
- Listed
Install
No one-command install. Set it up from its source.
Alternatives · Evals
- metr-task-standard192 stars · 37 forks94.339
What it is
UK AISI's framework for safety evaluations of large language models, with task-based scaffolding and solver pipelines.
When to use it
UK AISI's framework for safety evaluations of large language models, with task-based scaffolding and solver pipelines.
Notes
Curated evals entry. Verified 2026-05-27.