Sentiment Classification on Hotel
0.017Generalization Gap (Δ)IRM
Evaluation Results
| Method | Links | |
|---|---|---|
| IRMprotocol=leave-one-out zero-shot transfer, aggregation=mean over 8 seeds2026.05 | 0.017 | |
| DROprotocol=leave-one-out zero-shot transfer, aggregation=mean over 8 seeds2026.05 | 0.017 | |
| DANNprotocol=leave-one-out zero-shot transfer, aggregation=mean over 8 seeds2026.05 | 0.021 | |
| Fishprotocol=leave-one-out zero-shot transfer, aggregation=mean over 8 seeds2026.05 | 0.025 | |
| AGMprotocol=leave-one-out zero-shot transfer, aggregation=mean over 8 seeds2026.05 | 0.031 | |
| Mask-onlyprotocol=leave-one-out zero-shot transfer, aggregation=mean over 8 seeds2026.05 | 0.032 | |
| BERTprotocol=leave-one-out zero-shot transfer, aggregation=averaged across all source-target pairs2026.05 | 0.059 | |
| RoBERTaprotocol=leave-one-out zero-shot transfer, aggregation=averaged across all source-target pairs2026.05 | 0.06 |