CAA
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Behavior Selection | CAA (50 random samples) | Accuracy (coordinate-ais, pair)98 | 22 | |
| Hallucination Steering | CAA | Runtime19.5 | 13 | |
| Concept Alignment | CAA (test) | AIC75.18 | 12 | |
| Open-ended behavior generation | CAA | CoAIS Score5.33 | 10 | |
| Hallucination | CAA | Accuracy (pair)88 | 8 |
Showing 5 of 5 rows