{"schema":"hottub-news/story/v1","story":"st_y62pthofgqfwak2dayna","title":"GPT-6.1 Sol nearly matches the no-CoT performance of GPT-6 Astra","title_en":null,"outlets":1,"brief":null,"page":"https://hottub.news/stories/gpt-6-1-sol-nearly-matches-the-no-cot-performance-of-gpt-6-astra-y62pthofgq","count":1,"items":[{"id":"n_y62pthofgqfwak2dayna","seq":820146,"kind":"article","title":"GPT-6.1 Sol nearly matches the no-CoT performance of GPT-6 Astra","url":"https://www.lesswrong.com/posts/LqSZZAriGqgsGDQe3/gpt-6-1-sol-nearly-matches-the-no-cot-performance-of-gpt-6","summary":"Summary I ran GPT-6 Sol and GPT-6.1 Sol on the task suite from Think Fast . Surprisingly, 6.1 Sol performs substantially better than 6 Sol, almost matching the performance of GPT-6 Astra. The plot below gives a quick overview of the results: Measured by mean accuracy across the 27 tasks, GPT-6.1 Sol closes 80% (95% CI: 72–87%) of the gap between GPT-6 Sol and GPT-6 Astra, and is closer to Astra than to GPT-6 Sol on 24 of 27 tasks. The likely reason behind this gap is that…","published":"2026-09-30T03:57:43Z","seen":"2026-09-30T04:33:48Z","lang":"en","source":{"id":"kite-lesswrong-com-7072cb","name":"lesswrong.com","domain":"lesswrong.com","outlet":{"wikidata":"Q13980547","name":"Less Wrong","type":"website","designations":["group blog","internet forum"]}},"via":"kite-lesswrong-com-7072cb","topics":["ai"],"authors":["Rauno Arike"],"entities":[{"id":"Q141554688","name":"GPT-6 Astra","type":"other"}],"story":"st_y62pthofgqfwak2dayna","category":"science-technology","category_p":0.7699999809265137,"sentiment":"neutral","tone":0.2800000011920929,"political":0.09000000357627869}]}