New hybrid LLM approach optimizes query augmentation for information retrieval

A new research paper introduces a hybrid approach to query augmentation for information retrieval, merging prompting-based and reinforcement learning (RL) methods. The study found that simple, training-free query augmentation often performs comparably to or better than more computationally expensive RL-based methods, especially with powerful large language models (LLMs). The proposed On-policy Pseudo-document Query Expansion (OPQE) method generates a pseudo-document to maximize retrieval performance, outperforming standalone prompting and RL-based rewriting. AI

IMPACT This research could lead to more efficient and effective information retrieval systems by optimizing LLM-based query augmentation.

RANK_REASON The cluster contains a research paper detailing a new method for query augmentation in information retrieval. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• information retrieval
• large language models
• On-policy Pseudo-document Query Expansion
• reinforcement learning
• Zhichao Xu

• paper
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →