List your company
ResearchFormerly WizWand

Language Understanding on MMLU (Delta Accuracy)

37.28Delta AccuracyLMNet
-1.24168.759218.7628.7608May 19, 2025
Updated 5mo ago

Evaluation results

MethodLinks
LMNetBackbone=Qwen2.5-1.5B2025.0537.28
PromptBackbone=Qwen2.5-1.5B2025.0535.83
CODIBackbone=Qwen2.5-1.5B2025.0535.16
SFTBackbone=Qwen2.5-1.5B2025.0535.02
PredBackbone=Qwen2.5-1.5B2025.0534.75
COCONUTBackbone=Qwen2.5-1.5B2025.0533.85
LMNetBackbone=Llama3.2-1B-Instruct2025.0527.57
LMNetBackbone=Llama3.2-1B2025.0527.19
SFTBackbone=Llama3.2-1B-Instruct2025.0524.83
SFTBackbone=Llama3.2-1B2025.0524.65
PromptBackbone=Llama3.2-1B-Instruct2025.0524.3
LMNetBackbone=Qwen2.5-0.5B2025.0523.35
PromptBackbone=Qwen2.5-0.5B2025.0522.5
PredBackbone=Qwen2.5-0.5B2025.0522.36
SFTBackbone=Qwen2.5-0.5B2025.0521.5
CODIBackbone=Llama3.2-1B-Instruct2025.0521.3
PredBackbone=Llama3.2-1B-Instruct2025.0520.93
CODIBackbone=Llama3.2-1B2025.0520.69
CODIBackbone=Qwen2.5-0.5B2025.0519.48
COCONUTBackbone=Qwen2.5-0.5B2025.0519.22
COCONUTBackbone=Llama3.2-1B-Instruct2025.0518.54
LMNetBackbone=GPT2-XL2025.0513.1
PromptBackbone=Llama3.2-1B2025.0511.69
COCONUTBackbone=Llama3.2-1B2025.059.88
SFTBackbone=GPT2-XL2025.057.4
PredBackbone=Llama3.2-1B2025.056.05
CODIBackbone=GPT2-XL2025.055.01
COCONUTBackbone=GPT2-XL2025.053.63
PromptBackbone=GPT2-XL2025.051.69
PredBackbone=GPT2-XL2025.050.24