Long Text Generation on Stylized Feedback Generation Benchmark
24.2R-1PAT
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| PATLLM=Qwen32026.04 | 24.2 | 17.9 | 21 | 3.357 | |
| PATLLM=LlaMA32026.04 | 23.1 | 17.5 | 17.9 | 3.171 | |
| PGraphLLM=Qwen32026.04 | 21.3 | 15.3 | 19.1 | 3.685 | |
| PGraphLLM=LlaMA32026.04 | 21.1 | 15.2 | 17.5 | 3.29 | |
| GraSPeRLLM=Qwen32026.04 | 19.6 | 14.9 | 16 | 3.04 | |
| GraSPeRLLM=LlaMA32026.04 | 19.1 | 15.4 | 13.4 | 2.93 | |
| LaMPLLM=LlaMA32026.04 | 18.1 | 12.2 | 16.4 | 2.873 | |
| LaMPLLM=Qwen32026.04 | 17.9 | 12.4 | 16.1 | 3.107 |