| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Text Classification | Emotion (test) | Accuracy79.9 | 38 | |
| Text Classification | Emotion | ASR (%)0.808 | 36 | |
| Emotion Classification | Emotion | Accuracy87.68 | 27 | |
| Human Sensing | Emotion | ATR Ratio1.73 | 24 | |
| Explanation Faithfulness | Emotion | Delta AF Score5.637 | 24 | |
| Emotion Classification | Emotion | Delta 1 (Emotion)3.27 | 24 | |
| Text Classification | Emotion | Accuracy91.4 | 22 | |
| Emotion Classification | Emotion (Out-of-domain) | F1 Score0.4286 | 22 | |
| Classification | Emotion | ASR100 | 15 | |
| Emotion classification | Emotion | Correlation Coefficient0.6 | 12 | |
| OOD Detection | Emotion | AUROC0.997 | 12 | |
| Open-set selective classification | Emotion (test) | AUAC95.2 | 12 | |
| Membership Inference Attack | Emotion | TPR @ 0.1% FPR1.75 | 11 | |
| Steering | Emotion | Steering Success92.7 | 11 | |
| Emotion Editing | Emotion Easy Task | WER1.4 | 11 | |
| Multi-class Classification | Emotion (EM) | Accuracy54.58 | 11 | |
| Text Classification | Emotion multi-class (five random splits) | NLL0.152 | 9 | |
| Classification | Emotion 10-shot | Accuracy65.5 | 9 | |
| Classification | Emotion 5-shot | Accuracy56.3 | 9 | |
| Classification | Emotion 3-shot | Accuracy (3-shot)55.5 | 9 | |
| Text Classification | Emotion | Total Communication Time ($10^3$ s)2.25 | 9 | |
| Embedding Quality Evaluation | emotion dataset | Angular Distortion Index (Class)0.0451 | 8 | |
| Text Classification | Emotion | Accuracy89 | 8 | |
| Tree Explanation | emotion | Runtime (ms/instance)11.3 | 6 | |
| Text classification | emotion | Execution Time (ms)11.3 | 6 |