OpenAI chief scientist reveals internal perplexity numbers, proposes cross-lab limits

Antique_Atmosphere97 · reddit · 2026-11-23

OpenAI's chief scientist published 'Perplexity and Intelligence', internally disclosing perplexity figures: 0.91 in the GPT-4 era, now under 0.8 — and arguing frontier labs should adopt cross-lab limits based on this metric, an unusual move to anchor capability-governance discussions in a quantitative training measure.

Related event: OpenAI discloses pretraining perplexity data, proposes shared safety thresholds(27 posts)→

Original post →

More from Models

Models channel →