What AI says about CER (Combining Evidence and Reasoning)

CER is an automated fact-checking system for biomedical claims that uses PubMed as a knowledge base and a large language model to assess the veracity of medical statements. It generates justifications for its assessments and classifies claims with zero-shot and fine-tuned models, with fine-tuning improving robustness across diverse datasets. The repository represents PRAISELab-PicusLab’s research on combining evidence and reasoning for biomedical fact-checking.