Technology ❯ AI Development ❯ Model Training ❯ Limitations of LLMs
The unreviewed preprint introduces a taxonomy that formalizes privacy and integrity threats in retrieval‑augmented systems.
