19 lines
1.6 KiB
Markdown
19 lines
1.6 KiB
Markdown
|
---
|
||
|
|
license: cc-by-sa-4.0
|
||
|
|
pipeline_tag: question-answering
|
||
|
|
library_name: transformers
|
||
|
|
base_model:
|
||
|
|
- Qwen/Qwen2.5-3B
|
||
|
|
---
|
||
|
|
|
||
|
|
# Search and Refine During Think: Autonomous Retrieval-Augmented Reasoning of LLMs
|
||
|
|
|
||
|
|
The model is presented in the paper [Search and Refine During Think: Autonomous Retrieval-Augmented Reasoning of LLMs](https://huggingface.co/papers/2505.11277).
|
||
|
|
|
||
|
|
# Paper abstract
|
||
|
|
|
||
|
|
Large language models have demonstrated impressive reasoning capabilities but are inherently limited by their knowledge reservoir. Retrieval-augmented reasoning mitigates this limitation by allowing LLMs to query external resources, but existing methods often retrieve irrelevant or noisy information, hindering accurate reasoning. In this paper, we propose AutoRefine, a reinforcement learning post-training framework that adopts a new ``search-and-refine-during-think'' paradigm. AutoRefine introduces explicit knowledge refinement steps between successive search calls, enabling the model to iteratively filter, distill, and organize evidence before generating an answer. Furthermore, we incorporate tailored retrieval-specific rewards alongside answer correctness rewards using group relative policy optimization. Experiments on single-hop and multi-hop QA benchmarks demonstrate that AutoRefine significantly outperforms existing approaches, particularly in complex, multi-hop reasoning scenarios. Detailed analysis shows that AutoRefine issues frequent, higher-quality searches and synthesizes evidence effectively.
|
||
|
|
|
||
|
|
# Code
|
||
|
|
|
||
|
|
The code for this project is available on GitHub: [https://github.com/syr-cn/AutoRefine](https://github.com/syr-cn/AutoRefine)
|