The harness optimizes GPT-5.5's environment to outperform Claude Fable.
The harness optimizes GPT-5.5's environment to outperform Claude Fable.
@jeffclune @lilianweng Lilian always brings the cleanest lit reviews. Interesting to see ADAS and AI Scientist grouped as co-evolving rather than competing.
This blog is a great review of the current environment on self-improvement We leveraged a number of things here for our harness of harnesses, Zenith, which takes normal models above Fable level on the hardest tasks https://ii.inc/blog/post/zenith https://twitter.com/lilianweng/status/2074372369213428144
Nice to see Automated Design of Agentic Systems (ADAS), the Darwin Gödel Machine, DGM-HyperAgents, and The AI Scientist all covered in this nice review of the growing field of ADAS by @lilianweng https://twitter.com/lilianweng/status/2074372369213428144
@lilianweng We implemented a number of these ideas here to take normal models to frontier level on frontierSWE as a pratical application https://ii.inc/blog/post/zenith
Better than any review paper https://twitter.com/lilianweng/status/2074372369213428144
Ask a question below.
Published answers will appear here.
