I am a university lecturer and group leader at the Chair of Statistical Learning and Data Science, Department of Statistics, LMU Munich. Currently, I serve as the teaching coordinator for the Munich Center for Machine Learning (MCML) and lead the “Methods Beyond Supervised Learning” focus group at LMU Munich’s Statistics Department.

Before joining LMU, I was a research fellow at Harvard Medical School and Mass General Hospital in Boston, working with Hiroyuki Yoshida’s group. During that time, I focused on probabilistic deep learning and robust representation learning for clinical data analysis. I earned my Ph.D. from the Hasso Plattner Institute (HPI) at Potsdam University, where I conducted research on robust deep learning models for medical image analysis under the supervision of Prof. Christoph Meinel. I also hold an M.Sc. in Artificial Intelligence from the Department of Computer Science, Shiraz University, and a Bachelor’s degree in Computer Science, Software Engineering.

Research

My research interests include both technical and theoretical skills of machine learning with a focus on its transformative applications within the realm of computer vision, natural language processing (NLP), and healthcare:

• Learning with minimal supervision (semi-supervised learning, positive-unlabeled learning, self-supervised learning)
• Probabilistic machine learning (scoring rule, Bayesian neural networks, uncertainty estimation)
• Deep metric learning (geometry learning, contrastive learning, ranking)
• Generative models (GANs, diffusion, flow models)

News

[2025.03] - My research proposal has been selected for the Amazon Research Award in the Responsible AI track, with a funding amount of $80,000.

[2025.03] - The Paper “Pre-trained molecular representations enable antimicrobial discovery” was accepted at Nature Biology.

[2025.01] - The Paper “Calibrating LLMs with Information-Theoretic Evidential Deep Learning” was accepted at ICLR 2025.

[2024.11] - The Paper “A Dual-Perspective Approach to Evaluating Feature Attribution Methods” was accepted at TMLR.

[2024.10] - Three papers accepted at the Medical Imaging Conference 2025

• Joint image clustering and self-supervised representation learning through debiased contrastive loss, Paper.
• Advancing reliability in self-supervised transformer models through hierarchical mask attention heads Paper.
• Feasibility assessment of multitasking in MRI neuroimaging analysis: tissue segmentation, cross-modality conversion and bias correction Paper.

[2024.07] - Paper entitled “Hyperbolic Contrastive Learning for Document Representations - A Multi-View Approach with Paragraph-level Similarities” accepted at ECAI 2024.

[2024.07] - Paper titled “Uncertainty-Aware Vision Transformers for Medical Image Analysis” accepted at the Uncertainty for Safe Utilization of Machine Learning in Medical Imaging workshop, part of MICCAI 2024.

[2024.05] - New preprint on Interpretable Layer Pruning for LLM. pdf.

[2024.05] - Two papers accepted at ECML-PKDD 2024:

• Diversified Ensemble of Independent Sub-Networks for Robust Self-Supervised Representation Learning pdf, code.
• Attention-Driven Dropout: A Simple Augmentation Method to Improve Self-supervised Contrastive Sentence Embeddings pdf, code.

[2024.04] - Paper accepted at Nature Biology on optimized model architectures for deep learning on genomic data pdf, code.

[2024.04] - New preprint on self-supervised molecular representations enables antimicrobial discovery. pdf.

[2024.03] - I will be a program committee member for the 2024 ECML-PKDD conference.

[2024.01] - Paper accepted at ICLR 2024 on probabilistic self-supervised learning. pdf, code

[2023.12] - Paper accepted at NeurIPS XMI 2024 on evaluation of feature explanation methods. pdf, code

[2023.10] - Paper accepted at ICMLA on uncertainty quantification of deep learning models for genomics sequences. pdf, code

[2023.08] - Paper accepted at Nature Communication Biology on a self-supervised model for genomics sequences (Self-GenomeNet). pdf, code

[2023.08] - Paper accepted at IEEE Conference on Computational Intelligence in Bioinformatics and Computational Biology (CIBCB) on neural architecture search for sequence genome. pdf, code

[2023.08] - Paper accepted in the Journal of Medical Image Analysis on Bayesian deep learning for medical image analysis. pdf, code

[2023.06] - Paper accepted in the Journal of Computer Vision and Image Understanding on self-supervised divergence learning. pdf, code

[2023.05] - Paper accepted at ACL 2023 on self-supervised geometry representation learning. pdf, code