Ok so the correct comparison would be to fix the harness on the old model and re... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

janalsncm 21 days ago | parent | context | favorite | on: GPT-6 Astra

Ok so the correct comparison would be to fix the harness on the old model and re-compare. Now they are comparing a new model to an old crippled one.

altcognito 21 days ago | [–]

They get 66% with the old harness, which is a lot better, but obviously not 100%

vlmutolo 21 days ago | [–]

They said 5.6 Sol got something around 40% with the corrected harness.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact