Intent Detection on Thomas
87.4Macro-F1Goal-based strategy
Evaluation Results
| Method | Links | |
|---|---|---|
| Goal-based strategyLLM=Mistral2026.03 | 87.4 | |
| Goal-based strategyLLM=Llama32026.03 | 86.2 | |
| Goal-based strategyLLM=Gemma2026.03 | 86.2 | |
| Zero-CoTLLM=Llama32026.03 | 82.5 | |
| Self-DebiasLLM=Llama32026.03 | 82.4 | |
| Self-DebiasLLM=Gemma2026.03 | 80.8 | |
| Zero-CoTLLM=Mistral2026.03 | 79.4 | |
| Zero-CoTLLM=Gemma2026.03 | 79.3 | |
| Self-DebiasLLM=Mistral2026.03 | 79.2 |