Text Classification on Tweet Offensive
86.03AccuracyBERT base + PiFi
Evaluation results
| Method | Links | ||
|---|---|---|---|
| BERT base + PiFiModel Backbone=BERT-base, PiFi Integration=true, Source LLM=Llama-3.1-8B2025.06 | 86.03 | 80.26 | |
| DeBERTa-V3 base + PiFiModel Backbone=DeBERTa-V3-base, PiFi Integration=true, Source LLM=Llama-3.1-8B2025.06 | 85.6 | 80.56 | |
| DeBERTa base + PiFiModel Backbone=DeBERTa-base, PiFi Integration=true, Source LLM=Llama-3.1-8B2025.06 | 85.47 | 80.65 | |
| RoBERTa base + PiFiModel Backbone=RoBERTa-base, PiFi Integration=true, Source LLM=Llama-3.1-8B2025.06 | 85.35 | 80.83 | |
| ELECTRA base + PiFiModel Backbone=ELECTRA-base, PiFi Integration=true, Source LLM=Llama-3.1-8B2025.06 | 84.99 | 78.48 | |
| RoBERTa baseModel Backbone=RoBERTa-base2025.06 | 83.6 | 76.82 | |
| DeBERTa-V3 baseModel Backbone=DeBERTa-V3-base2025.06 | 83.6 | 79.15 | |
| ELECTRA baseModel Backbone=ELECTRA-base2025.06 | 83.52 | 77.51 | |
| BERT baseModel Backbone=BERT-base2025.06 | 83.15 | 77.27 | |
| DeBERTa baseModel Backbone=DeBERTa-base2025.06 | 82.67 | 77.29 |