WHAT NEW IN X10.28 WED

AI collections from X · 22 collections · 58,925 posts

gpt-6POST

CVE benchmark: GPT-6 Astra leads at 90.6% pass@3

A model rediscovered 87.5% of benchmark CVEs at pass@3, tying DeepSeek V4 Pro for second behind GPT-6 Astra's 90.6%.

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
@teortaxesTexArchived 5 days agoOriginally posted 2026.10.22
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞) @teortaxesTexOriginally posted 2026.10.22
Open original on X

> - At pass@3, it rediscovered 87.5% of the benchmark CVEs (28/32). That ties DeepSeek V4 Pro 0813 for 2nd place, behind only GPT-6 Astra (90.6%) Honestly, crazy that 0813 already did that and no, they couldn't have distilled Astra for 0813, this is Whale Data Supremacy

This post quotes another post: “DeepSeek V4.1 Flash trails only GPT-6 Astra on a cybersecurity CVE benchmark”

x.com/teortaxestex/status/2113269217420091734Archived 2026.10.23 10:07