Small language-model-made edits to item descriptions can shift rankings in some RAG recommender configurations, but the paper's own demotion data point the wrong way and the detection-evasion claim is untested.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Stealthy LLM-Driven Data Poisoning Attacks Against Embedding-Based Retrieval-Augmented Recommender Systems
Small language-model-made edits to item descriptions can shift rankings in some RAG recommender configurations, but the paper's own demotion data point the wrong way and the detection-evasion claim is untested.