Sentiment Classification on aug-CEBaB-yelp
71.56AccuracyCLMN (roberta-base)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CLMN (roberta-base)Evaluation Level=Output-level (O)2025.10 | 71.56 | 81.16 | |
| roberta-baseEvaluation Level=Backbone2025.10 | 71.21 | 80.92 | |
| bert-base-uncasedEvaluation Level=Backbone2025.10 | 69.49 | 79.72 | |
| CLMN (bert-base-uncased)Evaluation Level=Output-level (O)2025.10 | 69.26 | 79.62 | |
| CLMN (bert-base-uncased)Evaluation Level=Reasoning-level (R)2025.10 | 65.35 | 76.49 | |
| CLMN (roberta-base)Evaluation Level=Reasoning-level (R)2025.10 | 64.68 | 76.51 | |
| CLMN (gpt2)Evaluation Level=Output-level (O)2025.10 | 63.6 | 76.39 | |
| gpt2Evaluation Level=Backbone2025.10 | 63.39 | 75.39 | |
| CLMN (gpt2)Evaluation Level=Reasoning-level (R)2025.10 | 62.72 | 75.76 | |
| lstmEvaluation Level=Backbone2025.10 | 47.54 | 65.65 | |
| CLMN (lstm)Evaluation Level=Output-level (O)2025.10 | 47.19 | 64.03 | |
| CLMN (lstm)Evaluation Level=Reasoning-level (R)2025.10 | 38.08 | 57.1 |