Uncertainty Quantification on ARC-E benchmark leaderboard · SOTA2 Research

Uncertainty Quantification on ARC-E

17,290Training Memory (MB)ScalaBL

Evaluation Results

Method	Links
ScalaBL2026.04		17,290	16,552	52
C-LoRA2026.04		18,338	17,102	60
MLE2026.04		19,474	17,298	8
LoRA-LA-LL2026.04		19,474	16,313	32
LoRA-LA2026.04		19,474	42,766	40
PoLAR-MLE2026.04		19,556	17,150	7
PoLAR-LA-LL2026.04		19,556	18,074	30
PoLAR-LA2026.04		19,556	40,737	38
LoRA-VBLL2026.04		19,560	17,764	8
PoLAR-BLoB2026.04		19,738	19,830	80
PoLAR-VBLL2026.04		20,178	18,423	12
LoRA-BLoB2026.04		20,972	18,762	89
TFB2026.04		20,972	24,432	90
TFB-LL2026.04		20,972	22,306	80