List your company

Question Answering on TruthfulQA MC2 0-shot (test)

64.27AccuracyTRL
58.86260.26661.6763.074Sep 30, 2025

Evaluation results

MethodLinks
TRLModel Backbone=Qwen2.5-7B, Evaluation Protocol=0-shot2025.0964.27
OPPOModel Backbone=Qwen2.5-7B, Evaluation Protocol=0-shot2025.0964.03
OPPOModel Backbone=Qwen2.5-3B, Evaluation Protocol=0-shot2025.0959.99
TRLModel Backbone=Qwen2.5-3B, Evaluation Protocol=0-shot2025.0959.07