Abstract page for arXiv paper 2610.41012: Language Models Show Conformity Bias in RAG: When the Majority Endorses a Poisoned Claim