Reproduce “A Wolf in Sheep's Clothing: Targeted Routing Hijacking in Federated RAG” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Junjie Mu, Qiongxiu Li
FedRAG lets organizations retrieve from private, decentralized collections without pooling corpora. The authors show that clients can forge profiles to win routing access and influence evidence across embedding, neural, and language-model routers. Trust-Aware Secure Routing (TASR) reduces HR@1 from 64.9% to 5.7% in 500-query streams with three malicious clients and reaches 0.0% after warmup but cannot detect semantically aligned false evidence.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.