Revision Generation on ArgRevision
41.38SARIFull-Finetuning
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Full-FinetuningModel=Qwen2.5-14B2026.01 | 41.38 | 60.96 | 25.35 | 42.56 | |
| Intention-TuningModel=Qwen2.5-14B2026.01 | 38.08 | 63.09 | 19.94 | 40.37 | |
| Full-FinetuningModel=Llama3.1-8B2026.01 | 37.8 | 62.79 | 15.63 | 38.74 | |
| IR-BaselineModel=Qwen2.5-14B2026.01 | 37.51 | 60.91 | 21.74 | 40.05 | |
| Intention-TuningModel=Llama3.1-8B2026.01 | 37.37 | 63.72 | 16.58 | 39.22 | |
| IR-BaselineModel=Llama3.1-8B2026.01 | 37.19 | 62.48 | 16.83 | 38.83 | |
| Intention-TuningModel=Mistral-7B2026.01 | 36.78 | 62.07 | 12.79 | 37.21 | |
| IR-BaselineModel=Mistral-7B2026.01 | 36.55 | 61.77 | 13.58 | 37.3 | |
| IST-BaselineModel=Qwen2.5-14B2026.01 | 36.22 | 63.81 | 14.29 | 38.11 | |
| Full-FinetuningModel=Mistral-7B2026.01 | 36 | 62.48 | 14.19 | 37.56 | |
| ICL-BaselineModel=Llama3.1-8B2026.01 | 35.66 | 52.24 | 19.08 | 35.66 | |
| LISA-BaselineModel=Llama3.1-8B2026.01 | 34.38 | 64.43 | 12.04 | 36.95 | |
| IST-BaselineModel=Llama3.1-8B2026.01 | 34.38 | 64.38 | 12.2 | 36.99 | |
| IST-BaselineModel=Mistral-7B2026.01 | 34.29 | 62.06 | 9.8 | 35.38 | |
| LISA-BaselineModel=Mistral-7B2026.01 | 34.14 | 62.23 | 11.28 | 35.88 | |
| LISA-BaselineModel=Qwen2.5-14B2026.01 | 34.14 | 63.65 | 13.4 | 37.06 | |
| ICL-BaselineModel=Qwen2.5-14B2026.01 | 33.82 | 55.9 | 15.02 | 34.91 | |
| CoT-BaselineModel=Llama3.1-8B2026.01 | 33.09 | 48.42 | 15.54 | 32.35 | |
| CoT-BaselineModel=Qwen2.5-14B2026.01 | 33.07 | 50.75 | 17.81 | 33.88 | |
| Copy-BaselineModel=N/A2026.01 | 30.07 | 80.6 | 3.75 | 38.14 | |
| ICL-BaselineModel=Mistral-7B2026.01 | 28.13 | 26.97 | 18.79 | 24.63 | |
| CoT-BaselineModel=Mistral-7B2026.01 | 24.36 | 11.24 | 22.07 | 19.22 | |
| BART-BaselineModel=N/A2026.01 | 16.78 | 2.34 | 15.31 | 11.48 |