Metraipolis › archive › dc_3a4dbbbe7acf
Tracing Query Expansion Effects through Sparse Autoencoder Features — arXiv:2609.06968
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.06968
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T15:05:22.432Z
- SHA-256 of the archived text
36c5ea7175a0852781225b1b8fdd90a877e25df500af268a04f04b2675a4a395- Stable URL of this page
https://metraipolis.com/documento/dc_3a4dbbbe7acf
How to cite
Tracing Query Expansion Effects through Sparse Autoencoder Features — arXiv:2609.06968 — arXiv (en), https://arxiv.org/abs/2609.06968 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3a4dbbbe7acf (Metraipolis archive dc_3a4dbbbe7acf, sha256 36c5ea7175a0852781225b1b8fdd90a877e25df500af268a04f04b2675a4a395)
BibTeX
@misc{dc3a4dbbbe7acf,
title = {Tracing Query Expansion Effects through Sparse Autoencoder Features — arXiv:2609.06968},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.06968},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3a4dbbbe7acf; sha256 36c5ea7175a0852781225b1b8fdd90a877e25df500af268a04f04b2675a4a395},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.06968. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Query expansion (QE) is a critical technique in information retrieval that enriches underspecified queries with additional textual context. However, its effect is often unreliable in modern dense retrieval, especially for strong off-the-shelf retrievers without retraining. Existing studies mainly examine expansion quality, semantic drift, or retrieval outcomes, but rarely explain how QE changes dense retrievers internally. In this work, we trace QE effects through sparse autoencoder (SAE) features. Using paired original and expanded queries, we decompose layer-wise retriever representations into sparse latent activations, identify QE-related latents from expansion-induced activation shifts, and interpret them with natural-language descriptions and retrieval cases. Our analysis shows that effective QE induces layer-concentrated changes in sparse latents aligned with retrieval intent and entity attributes, rather than only perturbing final query embeddings. SAE-based activation steering further validates these latents improve retrieval more consistently than random interventions or vanilla QE across four benchmarks, suggesting that SAEs can explain QE effects and offer a lightweight option for precise retrieval behavior modulation without query rewriting or retriever fine-tuning. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-09 in cs.IR). Categorie arXiv: Information Retrieval (cs.IR). Pagina del lavoro: https://arxiv.org/abs/2609.06968 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-16T14:49:33.948Z dalla pagina https://arxiv.org/catchup/cs.IR/2026-09-09?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).