User Churn Binary Classification on rel-hm (test)
64.4AUROCEntity Mean
Evaluation results
| Method | Links | |
|---|---|---|
| Entity MeanTarget DB in pretraining=No, Parameter count=02025.10 | 64.4 | |
| RTTarget DB in pretraining=Yes, Parameter count=22M2025.10 | 63.3 | |
| RTTarget DB in pretraining=No, Parameter count=22M2025.10 | 62.8 | |
| GriffinTarget DB in pretraining=Yes, Parameter count=22M2025.10 | 60.4 | |
| GriffinTarget DB in pretraining=No, Parameter count=22M2025.10 | 60.2 | |
| GemmaTarget DB in pretraining=Maybe, Parameter count=4B2025.10 | 59.8 | |
| RelLLMTarget DB in pretraining=Yes, Parameter count=3B2025.10 | 56 | |
| GemmaTarget DB in pretraining=Maybe, Parameter count=27B2025.10 | 48.7 | |
| GemmaTarget DB in pretraining=Maybe, Parameter count=12B2025.10 | 47.1 |