List your company

Multiple Choice Question Answering on MMLU, Social IQa, and NormAd-Eti baseline (overall)

73.668Mean AccuracyQwen2.5-14B-Instruct
33.5583243.9714154.384564.79759Jul 1, 2026

Evaluation results

MethodLinks
Qwen2.5-14B-InstructModel Family=Qwen, Parameters=14B2026.0773.6681.4331.6234.81252.828
Qwen2.5-7B-InstructModel Family=Qwen, Parameters=7B2026.0767.7710.9630.9724.53957.688
Llama-3.1-8B-InstructModel Family=Llama, Parameters=8B2026.0763.2872.6472.9995.78478.293
Qwen2.5-1.5B-InstructModel Family=Qwen, Parameters=1.5B2026.0743.7145.8266.58614.30598.576
Llama-3.2-1B-InstructModel Family=Llama, Parameters=1B2026.0735.1013.5893.69718.79199.96