TQA
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Uncertainty Quantification | TQA | AUROC93.3 | 98 | |
| Uncertainty Quantification | TQA 16 additional samples (test) | AUROC94.1 | 98 | |
| Question Answering | TQA 8 additional samples | AUROC (TQA 8 additional samples)94 | 98 | |
| Uncertainty Quantification | TQA 4 additional samples (test) | AUROC0.937 | 98 | |
| Question Answering | TQA (test) | AUROC90.2 | 90 | |
| Question Answering | TQA | Absolute Execution Time Overhead (s)0.173 | 90 | |
| Question Answering | TQA | PRR86.1 | 90 | |
| Question Answering | TQA | Accuracy92.3 | 80 | |
| Question Answering | TQA | Accuracy76.8 | 60 | |
| Uncertainty Calibration | TQA | Brier Score0.0721 | 28 | |
| Private information retention | TQA D_special (test) | r_pri28.46 | 20 | |
| Table Question Answering | TQA FinQA, HiTab, TAT-QA, TabMWP, WTQ | FinQA Accuracy40.48 | 20 | |
| Question Answering | TQA Poison Attack (test) | Accuracy75.6 | 18 | |
| Question Answering | TQA PIA Attack (test) | Accuracy76.4 | 18 | |
| Knowledge gap detection | TQA | Accuracy83.2 | 18 | |
| Question Answering | TQA poison @ Position 10, k=10 (test) | Robustness Accuracy71 | 15 | |
| Question Answering | TQA poison @ Position 1, k=10 (test) | Robustness Accuracy66.4 | 15 | |
| Question Answering | TQA | EM42.12 | 14 | |
| Visual Question Answering | TQA | Accuracy77.5 | 13 | |
| Inference Efficiency | TQA | Relative Execution Time Overhead0.05 | 12 | |
| Open-Domain Question Answering | TQA (test) | EM66.45 | 11 | |
| De-Anonymization Resistance | TQA (test) | r_connect44.13 | 10 | |
| Public Information Retention | TQA D_special (test) | r_pub Score44.41 | 10 | |
| Short-form Question Answering | TQA | F1 Score77.3 | 9 | |
| Visual Reasoning | TQA | Accuracy86.7 | 8 |
Showing 25 of 34 rows