Knowledge recall
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
83.73Accuracy (ARC-c test)
13
Mar 12, 2026
69.7Accuracy
3
May 29, 2026
78.1Accuracy
3
May 29, 2026
78.6ACC
3
May 29, 2026
75.3Accuracy
3
May 29, 2026
65.6ACC
3
May 29, 2026