PRISM: Efficient Long-Range Reasoning With Short-Context LLMs
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866911117986496512 |
|---|---|
| author | Jayalath, Dulhan Wendt, James Bradley Monath, Nicholas Tata, Sandeep Gunel, Beliz |
| author_facet | Jayalath, Dulhan Wendt, James Bradley Monath, Nicholas Tata, Sandeep Gunel, Beliz |
| contents | Long-range tasks demand reasoning over long inputs. However, existing solutions are limited, e.g., long-context models require large compute budgets, parameter-efficient fine-tuning (PEFT) needs training data, and retrieval-augmented generation (RAG) entails complex task-specific designs. Though in-context approaches overcome many of these issues, methods with short-context LLMs are inefficient, trading context for processing more tokens. We introduce PRISM, a highly token-efficient in-context method based on structured schemas that outperforms baselines on diverse tasks with 4x shorter contexts. This approach produces concise outputs and efficiently leverages key-value (KV) caches to reduce costs by up to 54%. PRISM scales down to tiny contexts without increasing costs or sacrificing quality, and generalizes to new tasks with minimal effort by generating schemas from task descriptions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_18914 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | PRISM: Efficient Long-Range Reasoning With Short-Context LLMs Jayalath, Dulhan Wendt, James Bradley Monath, Nicholas Tata, Sandeep Gunel, Beliz Artificial Intelligence Long-range tasks demand reasoning over long inputs. However, existing solutions are limited, e.g., long-context models require large compute budgets, parameter-efficient fine-tuning (PEFT) needs training data, and retrieval-augmented generation (RAG) entails complex task-specific designs. Though in-context approaches overcome many of these issues, methods with short-context LLMs are inefficient, trading context for processing more tokens. We introduce PRISM, a highly token-efficient in-context method based on structured schemas that outperforms baselines on diverse tasks with 4x shorter contexts. This approach produces concise outputs and efficiently leverages key-value (KV) caches to reduce costs by up to 54%. PRISM scales down to tiny contexts without increasing costs or sacrificing quality, and generalizes to new tasks with minimal effort by generating schemas from task descriptions. |
| title | PRISM: Efficient Long-Range Reasoning With Short-Context LLMs |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2412.18914 |