Skip to content
LiveNext 14:18:31
DevelopingSafety

OpenAI proposes perplexity as a shared metric for pacing frontier AI

OpenAI chief scientist Jakub Pachocki published its internal log-perplexity figures on Nov. 22: about 0.99 for GPT-3.5, 0.91 for GPT-4, and below 0.8 for current models. He proposes labs adopt shared evaluation norms and "common limits" built on perplexity, and notes the metric differs across datasets and ignores chain-of-thought gains.

HEAT
OUTLETS
2
NEW IN 6H
+2
FIRST SEEN
LAST SIGNAL

Primary source

OpenAIopenai.com/index/perplexity-and-intelligence

Coverage · 3 articles, oldest first

Saturday, November 21, 2026

Share
Share card: OpenAI proposes perplexity as a shared metric for pacing frontier AI

Share this article

Copying the image…

Open any post or message and paste it in. The link to TopFive is printed on the card.

Post it on

XLinkedInHacker NewsEmail
← The whole wire