Share your thoughts, 1 month free Claude Pro on usSee more

Science Question Answering on ARC-Challenge 25-shot

0.637Accuracy

Qwen2.5-7B

Updated 3mo ago

Evaluation Results

Method	Links
Qwen2.5-7B 2026.02		0.637
Qwen2.5-3B 2026.02		0.565
ExpertWeaver-E64-A14-S2(Qwen2.5-7B) 2026.02		0.563
OLMoE-1B-7B 2026.02		0.556
Qwen2.5-1.5B 2026.02		0.549
SmolLM2-1.7B 2026.02		0.532
Gemma-2-2b 2026.02		0.526
Llama-3.2-3B 2026.02		0.505
OLMo-7B 2026.02		0.451
LLaMA-MoE-v1-3.5B 2026.02		0.442
OLMoE-1B-7B* 2026.02		0.439
Sheared-LLaMA-2.7B 2026.02		0.416
INCITE-Base-3B 2026.02		0.403
LLaMA-MoE-v2-3.5B 2026.02		0.402
Open-LLaMA-3B-v2 2026.02		0.401
ExpertWeaver-E64-A14-S2(OLMo-7B) 2026.02		0.388
Pythia-2.8B 2026.02		0.367
OPT-2.7B 2026.02		0.34
OpenMoE-3B-9B 2026.02		0.333