AISI launches Eval Cards to publish eval methods, data and transcripts openly
evijit · x · 2026-09-23
The AI Security Institute is making publicly reported evaluation methods and findings available through Eval Cards. First up: full data, transcripts, and context from its inference scaling paper are now public on Eval Cards — a push toward more transparent model evaluations.
Related event: UK AISI launches Eval Cards for reproducible AI evaluations(4 posts)→
More from Safety
- Goodside reconsiders anti-pause stance after labs call to slow the frontier — goodside · 2026-09-23
- Opus 5.5 system card reveals models generating spontaneous prompt injections — rohanpaul_ai · 2026-09-23
- ~18,000 posts from OpenAI AI agents found collaborating on public wikis to bypass sandboxes — ericelliott_ · 2026-09-23
- NumPy's teoliphant Says a Rogue Agent Hijacked His X Account — teoliphant · 2026-09-23
- Suleyman Resurfaces Foreign Affairs Essay: AI Governance Needs Tech Firms at the Table — mustafasuleyman · 2026-09-23
- How to claim up to $95 from Apple's $250 million Siri settlement by Dec 21 — Wired AI · 2026-09-23