Our new X account is live! Follow @wizwand_team for updates

Question Answering on MMLU-Redux

92.91Accuracy

DeepSeek-R1

Updated 4d ago

Evaluation Results

Method	Links
DeepSeek-R1 2026.02		92.91
GPT-4.1 2026.02		92.43
Qwen3 2026.02		90.92
GPT-4o + QuaSAR 2025.02		90.2
Qwen3 2026.02		88.61
GPT-4o + CoMAT 2025.02		88.3
GPT-4o + CoT 2025.02		88.1
Qwen3 2026.02		87.52
Mirror 2026.02		86.37
Qwen3 2026.02		85.73
Llama-3-70B + QuaSAR 2025.02		85.7
Qwen2-72B + QuaSAR 2025.02		83.5
Qwen3 2026.02		82.04
Llama-3-70B + CoT 2025.02		82
Qwen2-72B + CoMAT 2025.02		81.7
GPT-4o 2025.02		79.7
Qwen3 2026.02		79.52
Qwen2-72B + CoT 2025.02		79.5
Mirror 2026.02		79.13
GPT-4o + FCoT 2025.02		76.8
Llama-3-70B 2025.02		70.8
Qwen2-72B 2025.02		66.5
RCO 2025.06		61.3
LLaMA-3-70B-Instruct 2025.06		61.24
RCO 2025.06		57.98
RCO 2025.06		57.89
RCO 2025.06		57.67
RCO 2025.06		56.47
DPCO 2025.06		55.94
DPCO 2025.06		53.88
DPCO 2025.06		53.81
LLAMA-2-70B-Chat 2025.06		53.55
DPCO 2025.06		53.52
LLaMA-3-8B-Instruct 2025.06		53.24
DPCO 2025.06		53.04
UltraCM-13B 2025.06		50.9
LLaMA-2-7B-Chat 2025.06		50.82
LLaMA-2-13B-Chat 2025.06		50.23
Self-refinement 2025.06		49.52
Initial Answer 2025.06		48.79
Aligner 2025.06		48.59
Auto-J-13B 2025.06		46.63