collaborators

5 papers

cs.LG2026

CAT-GS: Balanced Multimodal Learning via Calibrated Gating and Fusion Surgery

Mahir Shahriar Tamim, Sharjil Khan, Md. Samiul Alim +3

End-to-end training of multimodal neural networks often exhibits unstable neural dynamics characterized by three coupled failure modes that degrade learning: (i) modality imbalance…

cs.LG2026

Gradient Extrapolation-Based Policy Optimization

Ismam Nur Swapnil, Aranya Saha, Tanvir Ahmed Khan +2

Reinforcement learning is widely used to improve the reasoning ability of large language models, especially when answers can be automatically checked. Standard GRPO-style training…

cs.CL2025

GRPO++: Enhancing Dermatological Reasoning under Low Resource Settings

Ismam Nur Swapnil, Aranya Saha, Tanvir Ahmed Khan +1

Vision-Language Models (VLMs) show promise in medical image analysis, yet their capacity for structured reasoning in complex domains like dermatology is often limited by data scarc…

cs.CV2025

CLARIFY: A Specialist-Generalist Framework for Accurate and Lightweight Dermatological Visual Question Answering

Aranya Saha, Tanvir Ahmed Khan, Ismam Nur Swapnil +1

Vision-language models (VLMs) have shown significant potential for medical tasks; however, their general-purpose nature can limit specialized diagnostic accuracy, and their large s…

cs.AI2025

Compression Strategies for Efficient Multimodal LLMs in Medical Contexts

Tanvir A. Khan, Aranya Saha, Ismam N. Swapnil +1

Multimodal Large Language Models (MLLMs) hold huge potential for usage in the medical domain, but their computational costs necessitate efficient compression techniques. This paper…