@inproceedings{ranaldi-etal-2025-eliciting,
    title = "Eliciting Critical Reasoning in Retrieval-Augmented Generation via Contrastive Explanations",
    author = "Ranaldi, Leonardo  and
      Valentino, Marco  and
      Freitas, Andre",
    editor = "Chiruzzo, Luis  and
      Ritter, Alan  and
      Wang, Lu",
    booktitle = "Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)",
    month = apr,
    year = "2025",
    address = "Albuquerque, New Mexico",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/landing_page/2025.naacl-long.557/",
    pages = "11168--11183",
    ISBN = "979-8-89176-189-6",
    abstract = "Retrieval-augmented generation (RAG) have emerged as a critical mechanism in contemporary NLP to support Large Language Models (LLMs) in systematically accessing richer factual context. However, the integration of RAG mechanisms bring its inherent challenges, as LLMs need to integrate potentially noisy contexts. Recent studies have shown that LLMs still struggle to critically analyse RAG-based in-context information, a limitation that may lead to incorrect inferences and hallucinations. In this paper, we investigate how to elicit critical arguments in RAG via contrastive explanations. In particular, we propose Contrastive-RAG (CRAG), a framework that (i) retrieves relevant documents given a query,(ii) selects and exemplifies relevant passages, and (iii) generates explanations that explicitly contrast the relevance of the passages to (iv) support the final answer. We show the impact of C-RAG building contrastive reasoning demonstrations from LLMs to instruct smaller models for retrieval-augmented tasks. Extensive experiments demonstrate that CRAG improves state-of-the-art RAG models while (a) requiring significantly fewer prompts and demonstrations and (b) being robust to perturbations in the retrieved documents."
}