List your company

Multiple choice question answering on BT (test)

39.6AccuracyBase
35.12836.28937.4538.611May 26, 2026

Evaluation results

MethodLinks
BaseBackbone=QWEN2-7B2026.0539.6
PPT (Mean)Backbone=QWEN2-7B2026.0539.4
PPT (Distribution)Backbone=LLAMA3-8B2026.0539.3
PPT (Distribution)Backbone=QWEN2-7B2026.0538.9
BaseBackbone=LLAMA3-8B2026.0537.8
PPT (Mean)Backbone=LLAMA3-8B2026.0536.8
VerbalizedBackbone=LLAMA3-8B2026.0535.3