Free Question Answering on Auto categorization context-free
10.6BLEU ScoreGCoT-decoding
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GCoT-decodingModel=Qwen2.5-14B, Decoding strategy=Multi-path2026.04 | 10.6 | 30.5 | |
| GCoT-decodingModel=Gemma-7B, Decoding strategy=Multi-path2026.04 | 8.9 | 24.6 | |
| GCoT-decoding + SpanAlignModel=Gemma-7B, Decoding strategy=Multi-path2026.04 | 8.8 | 23.3 | |
| GCoT-decoding + SpanAlignModel=Qwen2.5-14B, Decoding strategy=Multi-path2026.04 | 8.8 | 30.2 | |
| GreedyModel=Qwen2.5-14B, Decoding strategy=Single-path2026.04 | 8.5 | 29 | |
| Beam SearchModel=Qwen2.5-14B, Decoding strategy=Multi-path2026.04 | 8.1 | 28.4 | |
| CoT-decoding + Prompt-basedModel=Qwen2.5-14B, Decoding strategy=Multi-path2026.04 | 8 | 29 | |
| Self-consistency + Prompt-basedModel=Gemma-7B, Decoding strategy=Multi-path2026.04 | 7.4 | 20.3 | |
| GCoT-decodingModel=Llama-3.1-8B, Decoding strategy=Multi-path2026.04 | 6.8 | 20 | |
| Temperature samplingModel=Qwen2.5-14B, Decoding strategy=Single-path2026.04 | 6.6 | 27.9 | |
| Temperature samplingModel=Gemma-7B, Decoding strategy=Single-path2026.04 | 6 | 13.6 | |
| GreedyModel=Gemma-7B, Decoding strategy=Single-path2026.04 | 5.8 | 16.8 | |
| Top-k samplingModel=Qwen2.5-14B, Decoding strategy=Single-path2026.04 | 5.6 | 26 | |
| Beam SearchModel=Gemma-7B, Decoding strategy=Multi-path2026.04 | 5.3 | 15 | |
| Self-consistency + Prompt-basedModel=Qwen2.5-14B, Decoding strategy=Multi-path2026.04 | 5.3 | 29.8 | |
| GreedyModel=Llama-3.1-8B, Decoding strategy=Single-path2026.04 | 5.1 | 16 | |
| Temperature samplingModel=Llama-3.1-8B, Decoding strategy=Single-path2026.04 | 4.9 | 13.3 | |
| Beam SearchModel=Llama-3.1-8B, Decoding strategy=Multi-path2026.04 | 4.7 | 15.4 | |
| Top-k samplingModel=Llama-3.1-8B, Decoding strategy=Single-path2026.04 | 4.5 | 11.2 | |
| GCoT-decoding + SpanAlignModel=Llama-3.1-8B, Decoding strategy=Multi-path2026.04 | 4.5 | 14.7 | |
| Top-k samplingModel=Gemma-7B, Decoding strategy=Single-path2026.04 | 4.3 | 13.7 | |
| Self-consistency + Prompt-basedModel=Llama-3.1-8B, Decoding strategy=Multi-path2026.04 | 3.1 | 14.1 | |
| CoT-decoding + Prompt-basedModel=Llama-3.1-8B, Decoding strategy=Multi-path2026.04 | 2 | 15.7 | |
| CoT-decoding + Prompt-basedModel=Gemma-7B, Decoding strategy=Multi-path2026.04 | 1.2 | 20.1 |