bias 1clinical language models 1debiasing 1fact verification 1narrative anchoring 1prompt engineering 1
From the 1 of 2 linked papers with an AI index.
2 papers
cs.CL2026
When Retrieval Helps and Distracts: Evaluating Evidence-Generating LLMs for Biomedical Claim Verification
Pritam Deka, Prabhjot Singh
Biomedical fact-checking systems must do more than predict whether a claim is supported, contradicted, or unaddressed: they should also produce evidence that is faithful, complete,…
cs.CL2026
Same Facts, Different Diagnosis: Measuring and Mitigating Narrative Anchoring in Clinical Language Models
Prabhjot Singh, Pritam Deka, Vijay Chennareddy
The paper identifies a bias called Narrative Anchoring, where clinical language models produce different diagnoses when the same medical facts are expressed in different sociolingu…