15 papers
Naming the Concepts Classifiers Rely On: Language-Anchored Decomposition for Faithful Explanation
Ahsan Habib Akash, Dipkamal Bhusal, Stacey Jones +3
Deep neural networks are widely deployed in high-stakes visual applications where interpretability is critical, yet existing explanations face a trade-off: post-hoc concept methods…
ProMoE-FL: Prototype-conditioned Mixture of Experts for Multimodal Federated Learning with Missing Modalities
Aavash Chhetri, Bibek Niroula, Eduard Vazquez +4
In this paper, we address the problem of multimodal federated learning with missing modality. Existing methods utilize an additional public dataset or perform naive feature synthes…
A Benchmark for Hallucination Detection in VLMs for Gastrointestinal Endoscopy
Aminu Lawal, Niyoj Oli, Sachin Acharya +3
Vision-language models (VLMs) are prone to hallucination, which remains a major barrier to their safe deployment in clinical practice. To date, most hallucination detection methods…
When Confidence Lacks Concepts: Interpretable OOD Detection via Representation Perturbations
Anju Chhetri, Pratik Shrestha, Ramesh Rana +3
Deep neural networks have achieved remarkable performance across medical imaging tasks, yet their tendency to overgeneralize under distributional shifts poses a major obstacle to s…
Two is better than one: A Collapse-free Multi-Reward RLIF Training Framework
Shourov Joarder, Diganta Sikdar, Ahsan Habib Akash +2
Reinforcement learning with verifiable rewards (RLVR) has substantially improved the reasoning ability of LLMs, but often depends on external supervision from human annotations or…
Investigating Trustworthiness of Nonparametric Deep Survival Models for Alzheimer's Disease Progression Analysis
Jacob Thrasher, Kaitlyn Heintzelman, Peter Martone +4
Alzheimer's Dementia (AD) is a progressive neurodegenerative disease marked by irreversible decline, making reliable modeling of its progression essential for effective patient car…