Trusted Source Alignment in Large Language Models
원문 초록
Large language models (LLMs) are trained on web-scale corpora that inevitably include contradictory factual information from sources of varying reliability. In this paper, we propose measuring an LLM property called trusted source alignment (TSA): the model's propensity to align with content produced by trusted publishers in the face of uncertainty or controversy. We present FactCheckQA, a TSA evaluation dataset based on a corpus of fact checking articles. We describe a simple protocol for evaluating TSA and offer a detailed analysis of design considerations including response extraction, claim contextualization, and bias in prompt formulation. Applying the protocol to PaLM-2, we find that as we scale up the model size, the model performance on FactCheckQA improves from near-random to up to 80% balanced accuracy in aligning with trusted sources.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
From Generation to Detection: A Multimodal Multi-Task Dataset for Benchmarking Health Misinformation
Communication Styles and Reader Preferences of LLM and Human Experts in Explaining Health Information
Computational Fact-Checking of Online Discourse: Scoring scientific accuracy in climate change related news articles
Can LLMs Automate Fact-Checking Article Writing?
Fuzzy Deep Hybrid Network for Fake News Detection
AI Wizards at CheckThat! 2025: Enhancing Transformer-Based Embeddings with Sentiment for Subjectivity Detection in News Articles
토론
이 논문과 연결된 포럼 글이 아직 없습니다.