Multi-label biomedical classification on PubMed (test)
73.54Macro-F1EvoPool
Evaluation results
| Method | Links | |
|---|---|---|
| EvoPoolBackbone=Llama-3.1-8B2026.06 | 73.54 | |
| EvoPoolBackbone=RoBERTa-large2026.06 | 73.46 | |
| EvoPoolBackbone=Qwen2-1.5B2026.06 | 72.65 | |
| EvoPoolDownstream Model=RoBERTa-large2026.06 | 71.26 | |
| EvoPoolDownstream Model=Llama-3.1-8B2026.06 | 71.1 | |
| EvoPoolDownstream Model=Qwen3-1.7B2026.06 | 70.95 | |
| EvoPoolBackbone=gpt-4o-mini2026.06 | 70.26 | |
| Self-TrainDownstream Model=RoBERTa-large2026.06 | 69.75 | |
| LLM-AnnotationBackbone=Llama-3.1-8B2026.06 | 67.27 | |
| LLM-AnnotationBackbone=Qwen2-1.5B2026.06 | 66.22 | |
| LLM-AnnotationBackbone=RoBERTa-large2026.06 | 65.61 | |
| GoldenDownstream Model=RoBERTa-large2026.06 | 62.29 | |
| AlchemistBackbone=Llama-3.1-8B2026.06 | 56.84 | |
| Self-TrainDownstream Model=Llama-3.1-8B2026.06 | 56.18 | |
| GoldenDownstream Model=Llama-3.1-8B2026.06 | 55.93 | |
| DataSculptBackbone=Llama-3.1-8B2026.06 | 54.31 | |
| AlchemistBackbone=RoBERTa-large2026.06 | 49.67 | |
| DataSculptBackbone=Qwen2-1.5B2026.06 | 48.24 | |
| AlchemistBackbone=Qwen2-1.5B2026.06 | 47.06 | |
| Self-TrainDownstream Model=Qwen3-1.7B2026.06 | 45.66 | |
| GoldenDownstream Model=Qwen3-1.7B2026.06 | 44.77 | |
| LLM-AnnotationDownstream Model=RoBERTa-large2026.06 | 43.96 | |
| LLM annotationBackbone=gpt-4o-mini2026.06 | 43.75 | |
| LLM-EvalDownstream Model=RoBERTa-large2026.06 | 43.75 | |
| LLM-EvalDownstream Model=Qwen3-1.7B2026.06 | 43.75 | |
| LLM-EvalDownstream Model=Llama-3.1-8B2026.06 | 43.75 | |
| LLM-AnnotationDownstream Model=Llama-3.1-8B2026.06 | 43.53 | |
| LLM-AnnotationDownstream Model=Qwen3-1.7B2026.06 | 43.13 | |
| DataSculptBackbone=RoBERTa-large2026.06 | 37.74 | |
| AlchemistBackbone=gpt-4o-mini2026.06 | 14.61 | |
| AlchemistDownstream Model=RoBERTa-large2026.06 | 13.99 | |
| AlchemistDownstream Model=Llama-3.1-8B2026.06 | 13.97 | |
| AlchemistDownstream Model=Qwen3-1.7B2026.06 | 13.9 | |
| DataSculptBackbone=gpt-4o-mini2026.06 | 0.69 | |
| DataSculptDownstream Model=Llama-3.1-8B2026.06 | 0.53 | |
| DataSculptDownstream Model=Qwen3-1.7B2026.06 | 0.47 | |
| DataSculptDownstream Model=RoBERTa-large2026.06 | 0.43 |