STRUCTSURVEY: Structured Agentic Retrieval for Automated Survey Paper Generation

2026-07-03T08:52:17Zb6f4ee1bb94ab37b6b9d1c1f94de8fee13da253dfd53a77c59c2709372ee768b
HNSWRAGagentic-searchbenchmarksdata-exfiltrationembedding-inference-attackgraph-spannerskeystroke-biometricsknowledge-graphsmodel-fingerprintingoffice-file-processingpersonalizationprivacyreranker-evasionretrieval-augmented-generationstructured-retrieval

What happened

Collection of 12 recent ML/IR papers (arXiv 2026-07-03) focused on retrieval, RAG, personalization, and indexing. Notable security/privacy-relevant work: “Embedding Inference Attack” demonstrates a practical black‑box model‑fingerprinting attack that identifies which embedding model a closed API uses (even when only the unordered retrieved documents are observable) and can bypass rerankers; proposed mitigations include similarity thresholds. Related papers raise privacy and safety concerns from personalization and behavioral signals (keystroke dynamics, peer-profile completion, user-history–dr

Why it matters

A reviewed impact interpretation has not been published for this record.

Evidence and limitations

Source ID
arxiv_cs_ir
Record identifier
b6f4ee1bb94ab37b6b9d1c1f94de8fee13da253dfd53a77c59c2709372ee768b
Enrichment time
2026-07-03T08:52:17Z
AI-assisted enrichment
Yes

This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.