List your company
ResearchFormerly WizWand

High-stakes specialized classification on DDI (test)

61.12Macro-F1EvoPool
9.348822.789436.2349.6706Jun 1, 2026
Updated 5mo ago

Evaluation results

MethodLinks
EvoPoolBackbone=Llama-3.1-8B2026.0661.12
EvoPoolBackbone=RoBERTa-large2026.0659.74
Self-TrainDownstream Model=RoBERTa-large2026.0657.82
GoldenDownstream Model=RoBERTa-large2026.0657.74
EvoPoolBackbone=Qwen2-1.5B2026.0657.45
EvoPoolDownstream Model=Llama-3.1-8B2026.0653.79
EvoPoolDownstream Model=RoBERTa-large2026.0653.4
EvoPoolDownstream Model=Qwen3-1.7B2026.0652.93
Self-TrainDownstream Model=Qwen3-1.7B2026.0645.92
Self-TrainDownstream Model=Llama-3.1-8B2026.0645.81
LLM-AnnotationBackbone=Llama-3.1-8B2026.0645.49
EvoPoolBackbone=gpt-4o-mini2026.0645.39
GoldenDownstream Model=Qwen3-1.7B2026.0644.91
LLM-AnnotationBackbone=Qwen2-1.5B2026.0639.46
GoldenDownstream Model=Llama-3.1-8B2026.0638.63
DataSculptBackbone=Llama-3.1-8B2026.0634.63
DataSculptBackbone=Qwen2-1.5B2026.0629.55
AlchemistBackbone=Llama-3.1-8B2026.0627.16
LLM-AnnotationBackbone=RoBERTa-large2026.0627.07
AlchemistBackbone=RoBERTa-large2026.0626.87
AlchemistBackbone=Qwen2-1.5B2026.0626.71
LLM annotationBackbone=gpt-4o-mini2026.0623.45
LLM-EvalDownstream Model=RoBERTa-large2026.0623.45
LLM-EvalDownstream Model=Qwen3-1.7B2026.0623.45
LLM-EvalDownstream Model=Llama-3.1-8B2026.0623.45
LLM-AnnotationDownstream Model=Llama-3.1-8B2026.0623.05
LLM-AnnotationDownstream Model=RoBERTa-large2026.0622.34
LLM-AnnotationDownstream Model=Qwen3-1.7B2026.0622.06
AlchemistDownstream Model=Qwen3-1.7B2026.0621.55
AlchemistDownstream Model=Llama-3.1-8B2026.0621.21
AlchemistDownstream Model=RoBERTa-large2026.0618.06
DataSculptDownstream Model=Llama-3.1-8B2026.0617.78
DataSculptDownstream Model=Qwen3-1.7B2026.0617.56
DataSculptBackbone=RoBERTa-large2026.0617.46
DataSculptDownstream Model=RoBERTa-large2026.0617.46
AlchemistBackbone=gpt-4o-mini2026.0611.89
DataSculptBackbone=gpt-4o-mini2026.0611.34