New Science Benchmark Exposes Weakness of Frontier AI Models in Research

AItonomy Foundation, with researchers from 34 institutions including Harvard, MIT and Stanford, released Science's Last Exam (SLE), a benchmark of 91 end-to-end science research tasks. Eight frontier models performed poorly, with the best scoring only 67, and all scored zero on fully autonomous research.

2026-11-06 ~ 2026-11-06 · 3 related posts

• Science's Last Exam: best AI agent scores 37.5/100, all score 0 on end-to-end research — omarsar0 · 2026-11-06

2 near-duplicate retellings: anshulkundaje · AccBalanced