List your company
ResearchFormerly WizWand

High-stakes specialized classification on Claude9 (test)

50.31Macro F1LLM-Annotation
5.943617.461828.9840.4982Jun 1, 2026
Updated 5mo ago

Evaluation results

MethodLinks
LLM-AnnotationBackbone=RoBERTa-large2026.0650.31
LLM-AnnotationBackbone=Qwen2-1.5B2026.0648.9
LLM-AnnotationBackbone=Llama-3.1-8B2026.0646.6
EvoPoolBackbone=Qwen2-1.5B2026.0645
EvoPoolBackbone=Llama-3.1-8B2026.0642.59
EvoPoolDownstream Model=Llama-3.1-8B2026.0639.08
EvoPoolDownstream Model=Qwen3-1.7B2026.0638.27
EvoPoolBackbone=gpt-4o-mini2026.0636.45
LLM-AnnotationDownstream Model=Llama-3.1-8B2026.0636.39
LLM-AnnotationDownstream Model=Qwen3-1.7B2026.0635.79
EvoPoolBackbone=RoBERTa-large2026.0635.28
LLM-AnnotationDownstream Model=RoBERTa-large2026.0634.69
EvoPoolDownstream Model=RoBERTa-large2026.0634.13
LLM annotationBackbone=gpt-4o-mini2026.0633.2
LLM-EvalDownstream Model=RoBERTa-large2026.0633.2
LLM-EvalDownstream Model=Qwen3-1.7B2026.0633.2
LLM-EvalDownstream Model=Llama-3.1-8B2026.0633.2
AlchemistBackbone=Qwen2-1.5B2026.0632.4
AlchemistBackbone=Llama-3.1-8B2026.0625.26
AlchemistDownstream Model=Qwen3-1.7B2026.0622.07
AlchemistBackbone=RoBERTa-large2026.0621.6
GoldenDownstream Model=Llama-3.1-8B2026.0621.11
AlchemistDownstream Model=Llama-3.1-8B2026.0619.41
DataSculptDownstream Model=Qwen3-1.7B2026.0619.09
AlchemistDownstream Model=RoBERTa-large2026.0618.67
AlchemistBackbone=gpt-4o-mini2026.0616.33
DataSculptBackbone=Llama-3.1-8B2026.0616.1
DataSculptDownstream Model=Llama-3.1-8B2026.0615
DataSculptBackbone=gpt-4o-mini2026.0614.58
DataSculptBackbone=Qwen2-1.5B2026.0612.03
Self-TrainDownstream Model=Llama-3.1-8B2026.0611.31
DataSculptBackbone=RoBERTa-large2026.0610.57
GoldenDownstream Model=RoBERTa-large2026.0610.57
DataSculptDownstream Model=RoBERTa-large2026.0610.57
Self-TrainDownstream Model=RoBERTa-large2026.0610.57
GoldenDownstream Model=Qwen3-1.7B2026.0610.53
Self-TrainDownstream Model=Qwen3-1.7B2026.067.65