I am a PhD student in Computer Science at Heidelberg University, affiliated with the Institute of Computer Science, where I am supervised by Prof. Dr. Artur Andrzejak (2023–present). I obtained my M.Sc. in Computational Mathematics from Tianjin University (2019–2022), supervised by Prof. Dr. Huaming Wu, and my B.Sc. in Computational Mathematics from Qingdao University of Science and Technology (2015–2019).

My research lies at the intersection of Software Engineering and Artificial Intelligence, with a focus on trustworthy AI, interpretable machine learning, and large language models for code. Previously, I worked on efficient DNN offloading strategies. My work has been published in venues such as ICML, ICLR, IEEE Transactions on Services Computing, IEEE Transactions on Green Communications and Networking, and IEEE Consumer Electronics Magazine.

• Trustworthy and Interpretable Large Language Models — Mechanistic interpretability, knowledge editing, and reliability in Transformer models.

• AI for Software Engineering — Code generation, translation, automated debugging, and intelligent developer tools.

• Explainable Artificial Intelligence (XAI) — Methods for transparent and interpretable machine learning.

• Distributed and Edge AI Systems — Efficient DNN offloading and collaborative intelligence across local–edge–cloud environments.

May 2026 Our paper on SVD-based efficient interpretability for Transformers was accepted to ICML 2026 (Spotlight).

Mar 2026 Our paper on explainable AI-based feature selection for energy-harvesting IoT networks was accepted to the International Symposium on Communication Systems, Networks and Digital Signal Processing.

Feb 2026 Our paper on precise and interpretable editing of code knowledge in large language models was accepted to ICLR 2026.

Jun 2025 Our work on identifying decisive code snippets for LLM-based code inference was released as a preprint.

Jan 2024 Our paper on interpretable error correction for code-to-code translation was accepted to ICLR 2024.

May 2023 Our survey on DNN migration in IoT systems was published in IEEE Consumer Electronics Magazine.

Apr 2022 Two papers on DNN offloading and inference acceleration in local–edge–cloud collaborative environments were published in IEEE Transactions on Services Computing and IEEE Transactions on Green Communications and Networking. One of the papers (IEEE TSC) was listed among the Top 50 Most Popular Articles.

See my Google Scholar for a complete list of publications.

• SVD as a Fast Interpretability Method for Transformers

          Min Xue, Artur Andrzejak

          International Conference on Machine Learning (ICML), 2026 (Spotlight) [PDF] 

• Precise and Interpretable Editing of Code Knowledge in Large Language Models

          Min Xue, Nikolai Bolik, Lennart Stöpler, Erik Imgrund, Janik Schmid, Artur Andrzejak

          International Conference on Learning Representations (ICLR), 2026 [PDF]

• A Consistent Pattern for Identifying Decisive Code Snippets for LLM-Based Code Inference

          Min Xue, Artur Andrzejak

          Preprint, 2025 [PDF]

• An Interpretable Error Correction Method for Enhancing Code-to-Code Translation

          Min Xue, Artur Andrzejak, Marla Leuther

          International Conference on Learning Representations (ICLR), 2024 [PDF]

• DNN Migration in IoTs: Emerging Technologies, Current Challenges, and Open Research Directions

          Min Xue, Huaming Wu, Ruidong Li

          IEEE Consumer Electronics Magazine, vol. 12, no. 3, pp. 28–38, 2023 [PDF]

• DDPQN: An Efficient DNN Offloading Strategy in Local-Edge-Cloud Collaborative Environments

          Min Xue, Huaming Wu*, Guang Peng, Katinka Wolter

          IEEE Transactions on Services Computing, vol. 15, no. 2, pp. 640–655, 2022(Ranked among the 50 Most Popular Articles) [PDF] 

• EosDNN: An Efficient Offloading Scheme for DNN Inference Acceleration in Local-Edge-Cloud Collaborative Environments

          Min Xue, Huaming Wu, Ruidong Li, Minxian Xu, Pengfei Jiao

          IEEE Transactions on Green Communications and Networking, vol. 6, no. 1, pp. 248–264, 2022 [PDF]