Unverified50% confidenceBenchmarkExact time
Experimental results show that the strongest model reaches only 52.18% on the Ask-F1 metric.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/30/2026
First Seen
Valid until: 11/13/2026
Sources
Related Entities
Related Claims
UnverifiedEven the best-performing model achieves an evidence ID F1 of only around 0.36, reflecting that current large models still have significant room for improvement on cross-disciplinary reasoning tasks requiring precise evidence localization.71% similarUnverifiedAmong the 21 models evaluated on SpaceCast-Bench, the strongest model reaches only 58.0% accuracy, while human performance is 87.2%.67% similarUnverifiedIn the setting closest to reality (where unstated constraints must be actively discovered), the best-performing model configuration achieves an average FDE-Gym score of only 0.56.66% similarUnverifiedWhen asked directly, the accuracy of the decision models falls in the 50.0%–67.0% range.65% similarUnverifiedThe model showed the highest confidence (95%) on underdetermined problems (problems with no unique correct answer), exhibiting a reverse calibration phenomenon.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1114223API
curl https://kongchang.com/api/v1/knowledge/claims/1114223MCP
get_claim(id=1114223)