phenomenon · first seen 15 Oct 2026
membership and identity inference attack
Also: embedding leakage attack
Attacks that infer private information about items or users from learned embeddings or model outputs.
Mentions per week
PapersCodeDiscussion
Earliest sources
- Encoded but Exposed: Language Embeddings can Reveal More Than They Should in Recommender SystemsPapers · 15 Oct 2026 · Linda Beková, Rodrigo Alves
Latest sources
- Encoded but Exposed: Language Embeddings can Reveal More Than They Should in Recommender SystemsPapers · 15 Oct 2026 · Linda Beková, Rodrigo Alves