Fine-grained multi-label classification on ICLR gold-labeled (test)
74.24Jaccard SimilarityTADDLE
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| TADDLE2026.05 | 74.24 | 48.28 | 43.32 | 50.09 | |
| Longformer-base-40962026.05 | 71.11 | 17.78 | 6.98 | 17.76 | |
| BERT-base-uncased2026.05 | 66.67 | 9.3 | 3.98 | 7.44 | |
| DeepSeek V3.2few-shot=true2026.05 | 40.3 | 24.36 | 30.02 | 23.42 | |
| GLM 4.7few-shot=true2026.05 | 40.22 | 35.98 | 31.63 | 33.08 | |
| Qwen3.5-Plusfew-shot=true2026.05 | 39.54 | 33.52 | 30.54 | 29 | |
| Qwen3-30Bfew-shot=true2026.05 | 35.65 | 34.94 | 27.98 | 28.93 | |
| BigBird-RoBERTa-base2026.05 | 30.68 | 9.64 | 5.01 | 7.12 |