List your company
ResearchFormerly WizWand

Instruction Following on MMLU

64.31MMLU AccuracyFull FT
61.689262.369663.0563.7304Aug 27, 2025
Updated 6mo ago

Evaluation results

MethodLinks
Full FTCost=×1, Backbone=Llama 3.1-8B, Evaluation Protocol=Fine-tuning2025.0864.31
S^2-SAMBackbone=Llama 3.1-8B, Evaluation Protocol=fine-tuning2025.0863.73
Bi-LoRABackbone=Llama 3.1-8B, Evaluation Protocol=fine-tuning2025.0863.67
Flat-LoRACost=×1, Backbone=Llama 3.1-8B, Evaluation Protocol=Fine-tuning2025.0863.67
Bi-LoRACost=×1, Backbone=Llama 3.1-8B, Evaluation Protocol=Fine-tuning2025.0863.67
ROPCost=×1, Backbone=Llama 3.1-8B, Evaluation Protocol=Fine-tuning2025.0863.63
HiRABackbone=Llama 3.1-8B, Evaluation Protocol=fine-tuning2025.0863.62
LoRA-oBARCost=×1, Backbone=Llama 3.1-8B, Evaluation Protocol=Fine-tuning2025.0863.62
PiSSABackbone=Llama 3.1-8B, Evaluation Protocol=fine-tuning2025.0863.59
DoRABackbone=Llama 3.1-8B, Evaluation Protocol=fine-tuning2025.0863.58
RWP_LoRACost=×1, Backbone=Llama 3.1-8B, Evaluation Protocol=Fine-tuning2025.0863.5
DeLoRABackbone=Llama 3.1-8B, Evaluation Protocol=fine-tuning2025.0863.49
LoRA-SAMCost=×2, Backbone=Llama 3.1-8B, Evaluation Protocol=Fine-tuning2025.0863.46
LoRA-nBARCost=×1, Backbone=Llama 3.1-8B, Evaluation Protocol=Fine-tuning2025.0863.45
LoRA-LookSAMCost=×1.x, Backbone=Llama 3.1-8B, Evaluation Protocol=Fine-tuning2025.0863.45
WSAMBackbone=Llama 3.1-8B, Evaluation Protocol=fine-tuning2025.0863.42
RWP_fullCost=×1, Backbone=Llama 3.1-8B, Evaluation Protocol=Fine-tuning2025.0863.4
LoRABackbone=Llama 3.1-8B, Evaluation Protocol=fine-tuning2025.0863.38
LoRACost=×1, Backbone=Llama 3.1-8B, Evaluation Protocol=Fine-tuning2025.0863.38
LoRA-ESAMCost=×1.x, Backbone=Llama 3.1-8B, Evaluation Protocol=Fine-tuning2025.0861.79