NeuroscienceBench Released: Best AI Models Pass Only About 50%

A team has released NeuroscienceBench, the first in a new benchmark series for basic science, featuring 91 evaluations testing whether AI agents can reliably interpret neuroscience experimental data. The best models pass only about 50%, with future benchmarks planned for immunology and aging.

2026-11-10 ~ 2026-11-10 · 2 related posts

• NeuroscienceBench debuts: 91 evals test if AI agents can analyze real neuroscience data — kenbwork · 2026-11-10

1 near-duplicate retellings: kenbwork