paper-with-me

Text Classification 벤치마크

Text Classification on HateXplain

8개 결과 · ⬇ CSV · JSON

Accuracy (2 classes)

0.6588 0.7141 0.7693 0.8246 0.8798 2024-01 2026-11 Space-XLNet — 0.8798 (2024-01-30) XLNet — 0.816 (2024-01-30) Space-BERT — 0.811 (2024-01-30) BERT-base — 0.6588 (2024-01-30) Space-XLNet — 0.8798 (2024-01-30) XLNet — 0.816 (2024-01-30) Space-BERT — 0.811 (2024-01-30) BERT-base — 0.6588 (2024-01-30) Space-XLNet — 0.8798 (2024-01-30)
RankModel Accuracy (2 classes)F1 Macro PaperCodeYear
1 Space-XLNet 0.87980.8797 Breaking Free Transformer Models: Task-specific Context Attribution Promises Improved Generalizability Without Fine-tuning Pre-trained LLMs stepantita/space-model 2024
2 XLNet 0.81600.8156 Breaking Free Transformer Models: Task-specific Context Attribution Promises Improved Generalizability Without Fine-tuning Pre-trained LLMs stepantita/space-model 2024
3 Space-BERT 0.81100.8108 Breaking Free Transformer Models: Task-specific Context Attribution Promises Improved Generalizability Without Fine-tuning Pre-trained LLMs stepantita/space-model 2024
4 BERT-base 0.65880.6555 Breaking Free Transformer Models: Task-specific Context Attribution Promises Improved Generalizability Without Fine-tuning Pre-trained LLMs stepantita/space-model 2024
5 Space-XLNet 0.87980.8797 Breaking Free Transformer Models: Task-specific Context Attribution Promises Improved Generalizability Without Fine-tuning Pre-trained LLMs stepantita/space-model 2024
6 XLNet 0.81600.8156 Breaking Free Transformer Models: Task-specific Context Attribution Promises Improved Generalizability Without Fine-tuning Pre-trained LLMs stepantita/space-model 2024
7 Space-BERT 0.81100.8108 Breaking Free Transformer Models: Task-specific Context Attribution Promises Improved Generalizability Without Fine-tuning Pre-trained LLMs stepantita/space-model 2024
8 BERT-base 0.65880.6555 Breaking Free Transformer Models: Task-specific Context Attribution Promises Improved Generalizability Without Fine-tuning Pre-trained LLMs stepantita/space-model 2024
1–8 / 8 페이지당 10 20 50 100