Resume

Summary

I am a Ph.D. candidate in Computer Science and Engineering at the University of Connecticut, with research and engineering experience in LLM agents, generative AI, RAG, agentic control and safety, foundation-model adaptation, large-scale evaluation, and ML systems engineering.

Education

Ph.D., Computer Science and Engineering, University of Connecticut, expected August 2026.
B.S.E., Computer Science and Engineering, University of Connecticut, 2019.
B.A., Mathematics, University of Connecticut, 2019.

Experience

Graduate Research Assistant, Laboratory of Machine Learning and Health Informatics, May 2017 - Present.

Skills

Python, C++, PyTorch, LLM agents, RAG, generative AI, diffusion models, asynchronous CPU-GPU pipelines, large-scale evaluation harnesses, backend services, and experiment automation.

Publications

Collaborative Phenotype Inference from Comorbid Substance Use Disorders and Genotypes

Jin Lu, Jiangwen Sun, Xinyu Wang, Henry R. Kranzler, Joel Gelernter, and Jinbo Bi Proceedings of IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp. 392-397, 2017.

Inferring Phenotypes from Substance Use via Collaborative Matrix Completion

Jin Lu, Jiangwen Sun, Xinyu Wang, Henry R. Kranzler, Joel Gelernter, Jinbo Bi. Inferring phenotypes from substance use via collaborative matrix completion. BMC Systems Biology, 12, 15-27, 2018.

Fragment-based Deep Molecular Generation Using Hierarchical Chemical Graph Representation and Multi-resolution Graph Variational Autoencoder

Zhenxiang Gao, Xinyu Wang, Blake Blumenfeld Gaines, Xuetao Shi, Jinbo Bi, Minghu Song. Fragment-based deep molecular generation using hierarchical chemical graph representation and multi-resolution graph variational autoencoder. Molecular Informatics, 42(5), 2200215, 2023.

Polyhedron Attention Module: Learning Adaptive-Order Interactions

T. Zhu, F. Dou, Xinyu Wang, Jiangwen Lu, Jinbo Bi. Polyhedron attention module: learning adaptive-order interactions. Advances in Neural Information Processing Systems 36 (NeurIPS), 9213-9225, 2023.

Cross-platform Prediction of Depression Treatment Outcome Using Location Sensory Data on Smartphones

Soumyashree Sahoo, Chinmaey Shende, Md. Zakir Hossain, Parit Patel, Yushuo Niu, Xinyu Wang, Shweta Ware, Jinbo Bi, Jayesh Kamath, Alexander Russel, Dongjin Song, Qian Yang, Bing Wang. Cross-platform Prediction of Depression Treatment Outcome Using Location Sensory Data on Smartphones. arXiv:2503.07883, 2025.

Leveraging Partial SMILES Validation Scheme for Enhanced Drug Design in Reinforcement Learning Frameworks

Xinyu Wang, Jinbo Bi, Minghu Song. Leveraging Partial SMILES Validation Scheme for Enhanced Drug Design in Reinforcement Learning Frameworks. arXiv:2505.00530, 2025.

DIAL: Data-Free Interference-Aware Layer-Scaling via Diagnostic Composition for Task-Vector Model Merging

Xinyu Wang, Kangwei Deng, Fei Dou, Jinbo Bi, Jiangwen Lu. DIAL: Data-Free Interference-Aware Layer-Scaling via Diagnostic Composition for Task-Vector Model Merging. Manuscript under review, 2026.

SIGMA: Structure-Invariant Generative Molecular Alignment for Chemical Language Models via Autoregressive Contrastive Learning

Xinyu Wang, Fei Dou, Jinbo Bi, Minghu Song. SIGMA: Structure-Invariant Generative Molecular Alignment for Chemical Language Models via Autoregressive Contrastive Learning. arXiv:2603.25062, 2026.