Research map / Large language models / Retrieval and knowledge
Inference Scaling Law for Retrieval Augmented Generation
Shu Zhou, Yuxuan Ao, Yunyang Xuan, Xin Wang, Tao Fan, Hao Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| increase retriever size (110M to 4.8B) (ablation) | RAG with T5-11B generator · NQ, TriviaQA, WebQuestions, FEVER | F1 | Improved |
| increase generator size (OPT-1.3B to 30B) (ablation) | RAG with BERT-large retriever · NQ, TriviaQA, WebQuestions, FEVER | F1 | Improved |
| increase number of retrieved documents (1 to 100) (ablation) | RAG with OPT-30B · NQ, TriviaQA, WebQuestions, FEVER | F1 | Improved |
| increase context window size to very large values (ablation) | RAG with T5-11B · NQ, TriviaQA, WebQuestions, FEVER | F1 | Mixed |
| scale generator rather than retriever under a limited latency budget (ablation) | RAG resource allocation · TriviaQA | F1 | Improved |
| use a mid-sized retriever rather than the largest (ablation) | RAG resource allocation · TriviaQA | F1 | Improved |
Closest papers by meaning
- Inference Scaling for Long-Context Retrieval Augmented Generation (ICLR 2025 · similarity 0.97)
- RAGGED: Towards Informed Design of Scalable and Stable RAG Systems (ICML 2025 · similarity 0.92)
- Long-Context LLMs Meet RAG: Overcoming Challenges for Long Inputs in RAG (ICLR 2025 · similarity 0.92)
- Speculative RAG: Enhancing Retrieval Augmented Generation through Drafting (ICLR 2025 · similarity 0.92)
- Accelerating Inference of Retrieval-Augmented Generation via Sparse Context Selection (ICLR 2025 · similarity 0.91)
- RAPID: Long-Context Inference with Retrieval-Augmented Speculative Decoding (ICML 2025 · similarity 0.91)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
