Pith. sign in

REVIEW 9 cited by

Transform Before You Query: A Privacy-Preserving Approach for Vector Retrieval with Embedding Space Alignment

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2507.18518 v2 pith:E3FTK65P submitted 2025-07-24 cs.IR

Transform Before You Query: A Privacy-Preserving Approach for Vector Retrieval with Embedding Space Alignment

classification cs.IR
keywords querytextembeddingsretrievalsteertextbfvectorapproximate
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Vector Database (VDB) can efficiently index and search high-dimensional vector embeddings from unstructured data, crucially enabling fast semantic similarity search essential for modern AI applications like generative AI and recommendation systems. Since current VDB service providers predominantly use proprietary black-box models, users are forced to expose raw query text to them via API in exchange for the vector retrieval services. Consequently, if query text involves confidential records from finance or healthcare domains, this mechanism inevitably leads to critical leakage of user's sensitive information. To address this issue, we introduce STEER (\textbf{S}ecure \textbf{T}ransformed \textbf{E}mbedding v\textbf{E}ctor\textbf{ R}etrieval), a private vector retrieval framework that leverages the alignment relationship between the semantic spaces of different embedding models to derive approximate embeddings for the query text. STEER performs the retrieval using the approximate embeddings within the original VDB and requires no modifications to the server side. Our theoretical and experimental analyses demonstrate that STEER effectively safeguards query text privacy while maintaining the retrieval accuracy. Even though approximate embeddings are approximations of the embeddings from proprietary models, they still prevent the providers from recovering the query text through Embedding Inversion Attacks (EIAs). Extensive experimental results show that Recall@100 of STEER can basically achieve a decrease of less than 5\%. Furthermore, even when searching within a text corpus of millions of entries, STEER achieves a Recall@20 accuracy 20\% higher than current baselines.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 9 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. SHARD: cell-keyed residual splitting for alignment-resistant private dense retrieval

    cs.CR 2026-06 conditional novelty 7.0

    SHARD shards private embedding residuals into cell-local keyed groups to raise the anchor requirement for alignment attacks by a factor of C while preserving full-dimensional nDCG@10 via encrypted reranking.

  2. Not All Entities are Created Equal: A Dynamic Anonymization Framework for Privacy-Preserving RAG

    cs.CR 2026-03 unverdicted novelty 7.0

    TRIP-RAG dynamically anonymizes only high-risk entities in RAG knowledge bases via three context-aware metrics, achieving privacy comparable to full anonymization with under 35% recall drop and up to 56% better genera...

  3. SHARD: cell-keyed residual splitting for alignment-resistant private dense retrieval

    cs.CR 2026-06 conditional novelty 6.0

    Cell-keyed residual splitting recovers raw dense-retrieval nDCG while raising diffuse known-plaintext alignment cost roughly C-fold, without cryptographic document privacy or full unlinkability.

  4. SHARD: cell-keyed residual splitting for alignment-resistant private dense retrieval

    cs.CR 2026-06 unverdicted novelty 6.0

    SHARD introduces cell-keyed residual splitting that turns dense retrieval embeddings into revocable, renewable, unlinkable templates resistant to alignment attacks while preserving exact utility under CKKS reranking.

  5. Hybrid privacy-aware semantic search: SVD-truncated document geometry and CKKS-encrypted query reranking under a restricted threat model

    cs.CR 2026-06 unverdicted novelty 6.0

    A hybrid scheme combines SVD-truncated and rotated document vectors with CKKS-encrypted query reranking to preserve semantic search ranking quality at million-document scale under a restricted threat model while resis...

  6. Not All Entities are Created Equal: A Dynamic Anonymization Framework for Privacy-Preserving RAG

    cs.CR 2026-03 conditional novelty 6.0

    Context-aware selective entity anonymization for RAG matches full-anonymization privacy with <35% Recall@k drop and up to 56% better generation quality than baselines.

  7. Hybrid privacy-aware semantic search: SVD-truncated document geometry and CKKS-encrypted query reranking under a restricted threat model

    cs.CR 2026-06 conditional novelty 5.0

    SVD truncation plus a secret rotation on documents and CKKS on queries preserves sub-second retrieval quality while collapsing off-the-shelf inversion, with document protection failing under known-plaintext Procrustes...

  8. Hybrid privacy-aware semantic search: SVD-truncated document geometry and CKKS-encrypted query reranking under a restricted threat model

    cs.CR 2026-06 conditional novelty 5.0

    SVD truncation plus a secret rotation and CKKS query reranking can hide numerical query slots at sub-second latency on 1M docs, but document privacy is only empirical obfuscation that fails with about k known pairs.

  9. Hybrid privacy-aware semantic search: SVD-truncated document geometry and CKKS-encrypted query reranking under a restricted threat model

    cs.CR 2026-06 unverdicted novelty 5.0

    Hybrid privacy method for semantic search truncates and rotates document vectors geometrically while encrypting queries with CKKS, preserving retrieval quality on 1M-document corpora under a restricted threat model.