List your company

Question Answering on HotpotQA (EM, F1, Trigger rate (T))

42.6EMUncertain-Calibrate
111.822.633.4Apr 7, 2026

Evaluation results

MethodLinks
Uncertain-CalibrateScope=local2026.0442.652.767.4
Verbal-CalibrateScope=global2026.044252.861.6
DRAGIN2026.0434.648.487.6
Ret-All2026.043444.9100
ADARAGUE2026.0427.837.257
No-Ret2026.0423.431.9—
Base-UncTok2026.0422.533.13.5
FLARE2026.0421.835.299.2
Base-Verbal2026.0421.228.815.2
SR-7BBackbone=7B2026.044.216.657.8
SR-13BBackbone=13B2026.042.61642.4