Long-form Question Answering refinement on ELI5 (test)
0.0381Error RateEIR
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| EIRFeedback=Fine-grained (EIR)2024.07 | 0.0381 | 0.13 | 0.88 | 0.96 | 0.92 | |
| GenericFeedback=Coarse-grained (GENERIC)2024.07 | 0.0606 | 0.22 | 0.84 | 0.91 | 0.87 | |
| Zero-shotModel=LLaMA2-13B-chat, Prompting=Zero-shot2024.07 | 0.0961 | 0.27 | 0.74 | 0.89 | 0.81 | |
| ImproveFeedback=Coarse-grained (IMPROVE)2024.07 | 0.1005 | 0.36 | 0.75 | 0.86 | 0.8 | |
| BaselineType=Original dataset instances2024.07 | 0.2293 | 0.82 | — | — | — |