VerifAI: A Verifiable Open-Source Search Engine for Biomedical Question Answering

Collection of recent ML/IR arXiv papers (Apr 13, 2026) covering retrieval-augmented generation, verification, recommender-robustness, test-time adaptation, agent skill automation, and reasoning-based reranking. Highlights: VerifAI — an open-source biomedical RAG pipeline that decomposes answers into atomic claims and verifies them with a fine-tuned NLI verifier (claims fewer hallucinated citations, outperforms GPT-4 on HealthVer); DITaR — targeted rectification for fake orders in sequential recommender systems; UATTA — uncertainty-aware offline test-time adaptation for text-based person search