Hate Speech Detection on ETHOS-SO
89.2AccuracySimilar
Evaluation results
| Method | Links | |
|---|---|---|
| SimilarLLM=GPT3-c (6.7B)2023.01 | 89.2 | |
| SimilarLLM=GPT2-m (355M)2023.01 | 86.9 | |
| SimilarLLM=GPT3-d (175B)2023.01 | 86.9 | |
| SimilarLLM=GPT3-b (1.3B)2023.01 | 86.7 | |
| Latent Concept LearningLLM=GPT3-c (6.7B)2023.01 | 86.2 | |
| Latent Concept LearningBackbone=GPT2-large, Configuration=concept tokens as prefixes2023.01 | 86.2 | |
| Latent Concept LearningLLM=GPT3-d (175B)2023.01 | 85.1 | |
| Latent Concept LearningLLM=GPT3-b (1.3B)2023.01 | 84.6 | |
| Latent Concept LearningLLM=GPT3-a (350M)2023.01 | 84.4 | |
| SimilarLLM=GPT3-a (350M)2023.01 | 84.1 | |
| Latent Concept LearningLLM=GPT2-m (355M)2023.01 | 83.7 | |
| Latent Concept LearningLLM=GPT2-xl (1.5B)2023.01 | 83 | |
| SimilarLLM=GPT2-xl (1.5B)2023.01 | 82.8 | |
| Latent Concept LearningLLM=GPT2-l (774M)2023.01 | 82.5 | |
| Latent Concept LearningLLM=GPT2 (124M)2023.01 | 80.9 | |
| SimilarLLM=GPT2-l (774M)2023.01 | 80.9 | |
| SimilarLLM=GPT2 (124M)2023.01 | 77.7 | |
| UniformLLM=GPT3-c (6.7B)2023.01 | 77.5 | |
| UniformLLM=GPT3-d (175B)2023.01 | 77.2 | |
| UniformLLM=GPT2-m (355M)2023.01 | 76.3 | |
| UniformLLM=GPT3-b (1.3B)2023.01 | 71 | |
| UniformLLM=GPT2-xl (1.5B)2023.01 | 67.8 | |
| UniformLLM=GPT2-l (774M)2023.01 | 67.6 | |
| UniformLLM=GPT3-a (350M)2023.01 | 65.5 | |
| UniformLLM=GPT2 (124M)2023.01 | 64.1 |