Four new arXiv papers explore advancements in dense retrieval, a key component for large language models in information retrieval tasks. The first paper introduces a privacy-preserving method using learned deep hashing to efficiently filter candidate documents, reducing computational cost while maintaining retrieval quality. Another paper proposes AnchorQE, a training-free method that improves dense retrieval by optimizing how generated query expansions are integrated with original queries. A third paper, LLM-QL, leverages large language models through query likelihood maximization as an auxiliary task to enhance retriever performance. Finally, AdaWidth presents a query-adaptive approach to reduce embedding dimensions, evaluating fewer dimensions for queries that require them, thus improving efficiency without sacrificing accuracy. AI
IMPACT These advancements in dense retrieval could lead to more efficient and accurate information retrieval systems, improving the performance of LLMs in various applications.
RANK_REASON Cluster consists of multiple academic papers published on arXiv detailing new methods and models in dense retrieval.
- AdaWidth
- alphaXiv
- arXiv
- CatalyzeX
- Connected Papers
- DagsHub
- Gotit.pub
- Hugging Face
- Litmaps
- LLM-QL
- ScienceCast
- scite Smart Citations
- AnchorQE
- Beir
- differential privacy
- large-language models
- Lotte
- MS MARCO
- Qwen3 32B
- retrieval-augmented generation
- TREC DL
AI-generated summary · Google Gemini · from 6 sources. How we write summaries →
