Sage-Taiyi: A Metacognitive Architecture for Scientific Discovery
Sage-Taiyi led all eight metrics on the authors’ scientific ideation benchmark, scoring 54.0 against AI Scientist-v2’s 47.9.
Taiyi-Bench covers 265 tasks drawn from 50 papers across ten disciplines, testing historical rediscovery and open-ended ideation. The system reassesses evidence, assumptions and unresolved questions to adjust its research targets and reasoning methods. Five expert-guided collaborations connected its proposals to experimental and computational validation. Reported chemistry results included raising cobalt-catalyzed carbonylation yield from trace amounts to 84% and photocatalytic allylic amination yield from 13% to 69%.
HF Daily Papers' note
Taiyi-Bench covers 265 tasks drawn from 50 papers across ten disciplines, testing historical rediscovery and open-ended ideation. The system reassesses evidence, assumptions and unresolved questions to adjust its research targets and reasoning methods. Five expert-guided collaborations connected its proposals to experimental and computational validation. Reported chemistry results included raising cobalt-catalyzed carbonylation yield from trace amounts to 84% and photocatalytic allylic amination yield from 13% to 69%.
HF Daily Papers' note
score 5