List your company
ResearchFormerly WizWand

Language Understanding on MMLU

82.33RACortexDebate
23.954839.109954.26569.4201Jul 18, 2023Dec 19, 2023May 21, 2024Oct 22, 2024Mar 25, 2025Aug 26, 2025Jan 27, 2026
Updated 8mo ago

Evaluation results

MethodLinks
CortexDebateType=Ours2025.0782.33
PRDType=Full Debate2025.0777.33
RECONCILEType=Full Debate2025.0775
GDType=Part Debate2025.0774
ChatEvalType=Full Debate2025.0773
NDType=Part Debate2025.0771.67
MLDType=Full Debate2025.0771.33
MaVType=No Debate2025.0769.33
LLAMA 2Size=70B, Shots=5-shot2023.0768.9
LLAMA 1Size=65B, Shots=5-shot2023.0763.4
LLAMA 2Size=34B, Shots=5-shot2023.0762.6
KEELTraining=SFT, Evaluation Setting=5-shot2026.0162.5
Pre-LNTraining=SFT, Evaluation Setting=5-shot2026.0160
Qwen-7BModel Size=7B, Status=final released2023.0858.2
LLAMA 1Size=33B, Shots=5-shot2023.0757.8
FalconSize=40B, Shots=5-shot2023.0755.4
LLAMA 2Size=13B, Shots=5-shot2023.0754.8
Baichuan2-7BModel Size=7B2023.0854.2
InternLM-7BModel Size=7B2023.0851
Qwen-VLInitialization=Intermediate Qwen-7B checkpoint2023.0850.7
Qwen-7BModel Size=7B, Status=intermediate, Usage=LLM initialization for Qwen-VL2023.0849.9
ChatGLM2-6BModel Size=6B2023.0847.9
MPTSize=30B, Shots=5-shot2023.0746.9
LLAMA 1Size=13B, Shots=5-shot2023.0746.9
LLAMA2-7BModel Size=7B2023.0846.8
LLAMA 2Size=7B, Shots=5-shot2023.0745.3
Baichuan-7BModel Size=7B2023.0842.3
LLAMA 1Size=7B, Shots=5-shot2023.0735.1
LLAMA-7BModel Size=7B2023.0835.1
MPTSize=7B, Shots=5-shot2023.0726.8
FalconSize=7B, Shots=5-shot2023.0726.2