List your company

Multiple-choice Question Answering on CBT-Bench Task 1 Q&A

96Q&A AccuracyDeepSeek-R1-8B
73.69279.483585.27591.0665Jul 3, 2026

Evaluation results

MethodLinks
DeepSeek-R1-8BParameters=8B2026.0796
GPT-OSSParameters=20B2026.0789.86
Gemma3Parameters=12B2026.0780.45
MistralParameters=7B2026.0774.55