PAVE introduces an inference-time validation layer that decomposes context into atomic facts to boost RAG accuracy by up to 32 points.
March 24, 2026
Original Paper
PAVE: Premise-Aware Validation and Editing for Retrieval-Augmented LLMs
arXiv · 2603.20673
The Takeaway
This adds a much-needed 'check-then-commit' mechanism to RAG pipelines. By making answer commitment auditable at the premise level, it significantly reduces hallucinations and increases the reliability of evidence-grounded systems.
From the abstract
Retrieval-augmented language models can retrieve relevant evidence yet still commit to answers before explicitly checking whether the retrieved context supports the conclusion. We present PAVE (Premise-Grounded Answer Validation and Editing), an inference-time validation layer for evidence-grounded question answering. PAVE decomposes retrieved context into question-conditioned atomic facts, drafts an answer, scores how well that draft is supported by the extracted premises, and revises low-suppo