Author

Mehrdad Fazli

h-index 5 View on Semantic Scholar

We have 2 of 17 papers

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

#artificial intelligence Preprint Apr 2026

VeriSim: A Configurable Framework for Stress-Testing Medical AI Under Patient Communication Noise

VeriSim is introduced, a patient simulation framework that injects controllable noise along six clinically grounded communication dimensions while substantially preserving each patient's medical record and is released as an open-source framework for rigorous stress-testing of clinical AI.

Sina Mansouri, Mohit Marvania, Vibhavari Ashok Shihorkar et al. · 0 citations

Preprint Sep 2026

Does Playing it Safe Count as Faithfulness? Reassessing LVLM Hallucination Mitigation Methods

It is argued that hallucination mitigation should be evaluated as a faithfulness--informativeness--capability trade-off rather than through hallucination scores alone, because improvements on hallucination benchmarks do not reliably transfer to broader multimodal capabilities.

Mehrdad Fazli, Sina Mansouri, Mohit Marvania et al. · 0 citations