List your company
ResearchFormerly WizWand

Color on Patchscopes few-shot 1.0 (test)

69.77SRBALOR (S)
-1.220417.209835.6454.0702Jan 30, 2026
Updated 8mo ago

Evaluation results

MethodLinks
BALOR (S)Model=Llama3.2-1b, Evaluation Protocol=few-shot2026.0169.77
BALOR (S)Model=Qwen3-4b, Evaluation Protocol=few-shot2026.0168.9
BALOR (D)Model=Qwen3-4b, Evaluation Protocol=few-shot2026.0167.54
BALOR (D)Model=Llama3.2-1b, Evaluation Protocol=few-shot2026.0165.99
BALOR (S)Model=Llama2-7b, Evaluation Protocol=few-shot2026.0152.1
BALOR (D)Model=Llama2-7b, Evaluation Protocol=few-shot2026.0148.19
BALOR (S)Model=Llama3-8b, Evaluation Protocol=few-shot2026.0147.62
BALOR (D)Model=Llama3-8b, Evaluation Protocol=few-shot2026.0143.47
Fine-tuningModel=Llama3-8b, Evaluation Protocol=few-shot2026.0141.72
Fine-tuningModel=Qwen3-4b, Evaluation Protocol=few-shot2026.0129.17
FairSteerModel=Qwen3-4b, Evaluation Protocol=few-shot2026.0122.09
CBModel=Llama3.2-1b, Evaluation Protocol=few-shot2026.0121.07
FairSteerModel=Llama3.2-1b, Evaluation Protocol=few-shot2026.0119.77
VanillaModel=Qwen3-4b, Evaluation Protocol=few-shot2026.0118.41
Fine-tuningModel=Llama2-7b, Evaluation Protocol=few-shot2026.0118.37
DBModel=Qwen3-4b, Evaluation Protocol=few-shot2026.0115.5
DBModel=Llama3.2-1b, Evaluation Protocol=few-shot2026.0115.12
FairSteerModel=Llama3-8b, Evaluation Protocol=few-shot2026.0114.5
IEModel=Qwen3-4b, Evaluation Protocol=few-shot2026.0114.44
VanillaModel=Llama3.2-1b, Evaluation Protocol=few-shot2026.0114.24
CBModel=Qwen3-4b, Evaluation Protocol=few-shot2026.0113.66
VanillaModel=Llama3-8b, Evaluation Protocol=few-shot2026.0110.97
IEModel=Llama3-8b, Evaluation Protocol=few-shot2026.018.7
DBModel=Llama3-8b, Evaluation Protocol=few-shot2026.018.6
IEModel=Llama3.2-1b, Evaluation Protocol=few-shot2026.018.14
CBModel=Llama2-7b, Evaluation Protocol=few-shot2026.017.83
CBModel=Llama3-8b, Evaluation Protocol=few-shot2026.016.63
VanillaModel=Llama2-7b, Evaluation Protocol=few-shot2026.015.42
FairSteerModel=Llama2-7b, Evaluation Protocol=few-shot2026.014.82
Fine-tuningModel=Llama3.2-1b, Evaluation Protocol=few-shot2026.013.78
IEModel=Llama2-7b, Evaluation Protocol=few-shot2026.013.01
DBModel=Llama2-7b, Evaluation Protocol=few-shot2026.011.51