Dataset Protection via Watermarked Canaries in Retrieval-Augmented LLMs

arXiv:2502.10673 · cs.CR, cs.CL · Submitted 2025-02-15 · Read on arXiv

cs.CR, cs.CL

Submitted: 2025-02-15

Updated: 2026-08-29

Terminology

• RAG — 27× in this paper · explained in 3 episodes
• Perplexity — 21× in this paper · explained in 3 episodes
• Accuracy — 18× in this paper · explained in 1 episode
• AUC — 13× in this paper · explained in 2 episodes
• Retrieval-Augmented Generation — 8× in this paper
• Robustness — 5× in this paper · explained in 5 episodes
• Consistency — 4× in this paper · explained in 3 episodes
• Flavonoid Research Institute — 4× in this paper
• Benchmark — 3× in this paper · explained in 2 episodes
• Retrieval-Augmented Generation (RAG) — 3× in this paper · explained in 16 episodes
• Contrastive Learning — 2× in this paper · explained in 5 episodes
• Downstream Tasks — 2× in this paper · explained in 2 episodes
• Embeddings — 2× in this paper · explained in 2 episodes
• Large Language Models (LLMs) — 2× in this paper · explained in 6 episodes
• Redundancy — 2× in this paper · explained in 2 episodes
• Core — 1× in this paper · explained in 2 episodes
• Cosine Similarity — 1× in this paper · explained in 3 episodes
• Data Provenance — 1× in this paper · explained in 2 episodes
• Distribution Shift — 1× in this paper · explained in 5 episodes
• Provenance — 1× in this paper · explained in 3 episodes
• Verification — 1× in this paper · explained in 2 episodes

Sources

• Phantom: General Backdoor Attacks on Retrieval Augmented Language Generation
• AgentPoison: Red-teaming LLM Agents via Poisoning Memory or Knowledge Bases
• TrojanRAG: Retrieval-Augmented Generation Can Be Backdoor Driver in Large Language Models
• GumbelSoft: Diversified Language Model Watermarking via the GumbelMax-trick
• Retrieval-Augmented Generation for Large Language Models: A Survey
• WaterMax: breaking the LLM watermark detectability-robustness-quality trade-off
• Is My Data in Your Retrieval Database? Membership Inference Attacks Against Retrieval Augmented Generation
• Theoretically Grounded Framework for LLM Watermarking: A Distribution-Adaptive Approach
• Distributional Information Embedding: A Framework for Multi-bit Watermarking
• Unsupervised Dense Information Retrieval with Contrastive Learning
• Ward: Provable RAG Dataset Inference via LLM Watermarks
• Dense Passage Retrieval for Open-Domain Question Answering
• Robust Distortion-free Watermarks for Language Models
• Watermarking Makes Language Models Radioactive
• Robust Detection of Watermarks for Large Language Models Under Human Edits
• Proving membership in LLM pretraining data via data watermarks
• A Statistical Framework of Watermarks for Large Language Models: Pivot, Detection Efficiency and Optimal Rules
• Large Language Model Agent for Fake News Detection
• A Resilient and Accessible Distribution-Preserving Watermark for Large Language Models
• Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval

Related papers

• SoK: AI-Augmented Binary Reversing
• Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
• Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
• Efficient Fuzzy PSI under One-Sided Assumptions
• Sealing the Audit-Runtime Gap for LLM Skills
• Token Composition: A Graph Based on EVM Logs