Tabular Reasoning on Data-scarce Specialized Domains (FinQA, HiTab, MultiHiertt) (test)
20.4Accuracy (FinQA)PivotTrace
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| PivotTrace2026.06 | 20.4 | 64.8 | 25.6 | 36.9 | |
| CoE2026.06 | 20.1 | 60.3 | 23.5 | 34.6 | |
| Consistency2026.06 | 19.5 | 61.3 | 23.8 | 34.9 | |
| Random2026.06 | 19.4 | 58.2 | 23.2 | 33.6 | |
| Self-Certainty2026.06 | 19.3 | 59.5 | 23.2 | 34 | |
| Entropy2026.06 | 19.2 | 61.5 | 24.1 | 34.9 | |
| CoT-Kinetics2026.06 | 18.7 | 61.9 | 23.2 | 34.6 |