Intent Detection on Comment Dataset In-Distribution
66.47F1 Score (Validation)BOT-MOD
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| BOT-MODUser Model=Mistral-7B2026.05 | 66.47 | 73.18 | 53.14 | |
| BOT-MODUser Model=Llama-3.12026.05 | 62.4 | 64.97 | 56.79 | |
| zero-shot+User Model=Mistral-7B2026.05 | 61.77 | 66.92 | 51.24 | |
| self-consistencyUser Model=Mistral-7B2026.05 | 58.3 | 58.87 | 56.97 | |
| zero-shotUser Model=Mistral-7B2026.05 | 58 | 59.33 | 55.07 | |
| BOT-MODUser Model=Qwen32026.05 | 57.48 | 60.14 | 51.76 | |
| BERTUser Model=Mistral-7B2026.05 | 53.92 | 60.12 | 41.85 | |
| zero-shot+User Model=Llama-3.12026.05 | 51.1 | 51.02 | 51.29 | |
| CoTUser Model=Mistral-7B2026.05 | 50.94 | 53.41 | 45.65 | |
| zero-shot+User Model=Qwen32026.05 | 50.77 | 51.57 | 48.93 | |
| self-consistencyUser Model=Llama-3.12026.05 | 50.53 | 47.75 | 57.71 | |
| zero-shotUser Model=Llama-3.12026.05 | 50.14 | 47.95 | 55.63 | |
| self-refineUser Model=Mistral-7B2026.05 | 48.56 | 48.63 | 48.86 | |
| BERTUser Model=Llama-3.12026.05 | 47.46 | 55.81 | 32.5 | |
| zero-shotUser Model=Qwen32026.05 | 46.16 | 42.41 | 56.31 | |
| self-consistencyUser Model=Qwen32026.05 | 45.92 | 42.24 | 55.8 | |
| CoTUser Model=Llama-3.12026.05 | 43.27 | 38.61 | 56.57 | |
| self-refineUser Model=Qwen32026.05 | 43.03 | 44.81 | 39.27 | |
| BERTUser Model=Qwen32026.05 | 41.73 | 49.45 | 28.09 | |
| CoTUser Model=Qwen32026.05 | 39.63 | 38.93 | 41.33 | |
| self-refineUser Model=Llama-3.12026.05 | 39.38 | 45.15 | 28.71 |