Two approaches for fact retrieval and hallucination identification: one using advanced LLMs with chain-of-thought prompting and Google snippets, and another using traditional NLP techniques including semantic ranking and token-level extraction.
Jul 1, 2025
A hybrid system combining PubMed API retrieval with LLM-driven methods for biomedical question answering.
Jan 1, 2025
A multilayer AI model combining CNNs, LSTMs, and attention mechanisms to detect semantic relationships in natural language, evaluated in SemEval 2024 Task 1.
Jun 1, 2024
This paper presents a lexical feature-based approach to identify news genres and persuasive framing in online content, utilizing lexical weirdness for classification.
Jul 1, 2023