Papers with 16-point boost
Can’t Remember Details in Long Documents? You Need Some R&R (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Long-context large language models miss important information in the middle of context documents . a recent study shows that LLMs can be used for document-based QA tasks . |
| Approach: | They propose a prompt-based method called *reprompting* and *in-context retrieval* to alleviate this effect in document-based QA. |
| Outcome: | The proposed method improves QA accuracy on documents up to 80k tokens in length. |