Customer Segmentation on BI sub-agent dataset Segmentation queries
97.2Syntax Pass RateEARS
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| EARSAbstention Training=Enabled, Sub-metric=Syntax validity2026.06 | 97.2 | — | 4.9 | |
| BaselineAbstention Training=Disabled, Sub-metric=Syntax validity2026.06 | 92.7 | — | — | |
| BaselineAbstention Training=Disabled, Sub-metric=Semantic correctness2026.06 | — | 61.9 | — | |
| EARSAbstention Training=Enabled, Sub-metric=Semantic correctness2026.06 | — | 82.7 | 33.6 |