First Double-Blind Evaluation of Proprietary Model Runs on Gemini 2.5 Flash-Lite
sethlazar · x · 2026-08-28
AVERI, in collaboration with Google DeepMind, OpenMinedOrg, and MLCommons, announced the first double-blind evaluation of a proprietary language model. The test used Gemini 2.5 Flash-Lite with new prompts from the AILuminate safety benchmark, executed inside a secure enclave to ensure strong privacy guarantees for both developers and evaluators.
More from Safety
- APEX-Harm benchmark: script-based injection attacks easily compromise agents in shared workspaces — KAUST · 2026-10-16
- Residual-stream forecasts predict agent violations before they happen, USC study shows — USC · 2026-10-16
- Perplexity open-sources SPACE, its Firecracker microVM agent sandbox — perplexity_ai · 2026-10-16
- AI safety drama: risk research called a 'company effort,' team lead responds — logangraham · 2026-10-16
- Brave's Leo adds end-to-end encrypted chat with GLM via attested TEE — gnukeith · 2026-10-16
- Cisco's Antares AI model ranks risky files for security teams, scores 0.823 on source but 0.000 on binaries — shashib · 2026-10-16