Shanghai AI Lab's Sage-Taiyi tops AI co-scientist benchmark, lifts reaction yields to 84%
Shanghai-AI-Laboratory · hf · 2026-11-10
Shanghai AI Laboratory introduces Sage-Taiyi, a metacognition-inspired AI co-scientist architecture that explicitly monitors and revises its understanding of evidence, assumptions, and open questions to decide what and how to investigate next.
- On Taiyi-Bench (265 tasks from 50 papers across 10 disciplines), it ranks first on all eight metrics of historical rediscovery and open-ended ideation, scoring 54.0 overall vs 47.9 for the strongest baseline, AI Scientist-v2; it also performs well on two external ideation benchmarks.
- Five expert-guided collaborations connect proposals to validation in chemistry, materials science, and engineering: reaction-condition optimization raises cobalt-catalyzed carbonylation yield from trace amounts to 84% and photocatalytic allylic amination from 13% to 69%.
The results support metacognitive regulation as a viable route from scientific ideation to practical validation.
More from Research
- Uno paper: diffusion adapters on AR LLMs beat speculative decoding with lossless speedups — LucaAmb · 2026-11-10
- DeepMind paper: diffusion model factors 68-bit semiprimes, but compute costs scale exponentially — weballergy · 2026-11-10
- Archon Robotics unveils Anchor 1, a whole-body humanoid controller with lowest tracking error — Scobleizer · 2026-11-10
- Alex Kendall's Team Releases Three-Part Series on Safety and Validation in End-to-End Learning — alexgkendall · 2026-11-10
- LLMs don't report injected pain states: decodable in probes, silent in speech, alters behavior — Emilio Ferrara · 2026-11-10
- TuF Labs tops ARC-AGI-2 leaderboard: general approaches now beat domain-specific tricks — rbhar90 · 2026-11-10