We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Information Retrieval

Title: Interpreting Dense Retrieval as Mixture of Topics

Abstract: Dense Retrieval (DR) reaches state-of-the-art results in first-stage retrieval, but little is known about the mechanisms that contribute to its success. Therefore, in this work, we conduct an interpretation study of recently proposed DR models. Specifically, we first discretize the embeddings output by the document and query encoders. Based on the discrete representations, we analyze the attribution of input tokens. Both qualitative and quantitative experiments are carried out on public test collections. Results suggest that DR models pay attention to different aspects of input and extract various high-level topic representations. Therefore, we can regard the representations learned by DR models as a mixture of high-level topics.
Subjects: Information Retrieval (cs.IR)
Cite as: arXiv:2111.13957 [cs.IR]
  (or arXiv:2111.13957v1 [cs.IR] for this version)

Submission history

From: Jingtao Zhan [view email]
[v1] Sat, 27 Nov 2021 18:35:44 GMT (986kb,D)

Link back to: arXiv, form interface, contact.