Publications: 1. Going Beyond Conventional OOD Detection (2024); 2. AdaSCALE: Adaptive Scaling for OOD Detection (2025); 3. ReweightOOD: Loss Reweighting for Distance-based OOD Detection (CVPR Workshops, 2024); 4. T2FNorm: Train-time Feature Normalization for OOD Detection in Image Classification (CVPR Workshops, 2024). Conference Reviewer: CVPR, MICCAI, ICCV, NeurIPS.
Research Experience
Projects: Developing solutions to enhance the reliability and safety of deep learning models for real-world deployment; Recently, keen interest in the safety and alignment aspects of foundation models.
Education
PhD Student at Dartmouth College, advised by Professor Andrew Campbell; Bachelor's degree from Department of Electronics and Computer Engineering, Pulchowk Campus, Tribhuvan University in 2022.
Background
Research Interests: Trustworthy machine learning; Focus: Enhancing the reliability and safety of deep learning models in safety-critical applications.