List your company

Selective Question Answering on AbstainQA (test)

13AccuracyEvoGM
-0.2083.2216.6510.079May 28, 2026

Evaluation results

MethodLinks
EvoGMBase model=Qwen2.5-1.5B2026.0513
Single BestBase model=Qwen2.5-1.5B2026.0511.9
BaseBase model=Qwen2.5-1.5B2026.0510.1
CMABase model=Qwen2.5-1.5B2026.057.8
Model SwarmBase model=Qwen2.5-1.5B2026.057.1
Task ArithmeticBase model=Qwen2.5-1.5B2026.056
PSO-MergingBase model=Qwen2.5-1.5B2026.055.9
Model SoupBase model=Qwen2.5-1.5B2026.054.8
DAREBase model=Qwen2.5-1.5B2026.054.3
TIESBase model=Qwen2.5-1.5B2026.052.8
MTLBase model=Qwen2.5-1.5B2026.050.3