I am a 3rd-year Ph.D. student in Computer Science at University of Southern California (USC), advised by Prof. Viktor Prasanna. Before that, I got my B.S. degree from Central South University (CSU), advised by Prof. Ning Gui. My current research interests lie in coding agents and LLM reasoning over graph-structured data.
Gangda Deng, Hongkuan Zhou, Rajgopal Kannan, Viktor Prasanna Mixture of Scope Experts at Test: Generalizing Deeper Graph Neural Networks with Shallow Variants. Code NeurIPS 2025 (3 Clear Accepts)
Zhaoling Chen*, Xiangru Tang*, Gangda Deng*, Fang Wu, Jialong Wu, Zhiwei Jiang, Viktor Prasanna, Arman Cohan, Xingyao Wang LocAgent: Graph-Guided LLM Agents for Code Localization. Code Slides ACL 2025 (2 Clear Accepts)
Gangda Deng*, Hongkuan Zhou*, Hanqing Zeng, Yinglong Xia, Christopher Leung, Jianbo Li, Rajgopal Kannan, Viktor Prasanna TASER: Temporal Adaptive Sampling for Fast and Accurate Dynamic Graph Representation Learning. Code Slides IPDPS 2024 (2 Strong Accepts, 1 Clear Accept)
Gangda Deng*, Yuxin Yang*, Ömer Faruk Akgül*, Hanqing Zeng, Yinglong Xia, Rajgopal Kannan, Viktor Prasanna Training Diverse Graph Experts for Ensembles: A Systematic Empirical Study. Code Arxiv 2025
Yuhong Liu*, Gangda Deng*, Hongkuan Zhou, Cauligi S. Raghavendra, Rajgopal Kannan, Ananthram Swami, Viktor Prasanna AP Selection in Uplink Cell-Free Massive MIMO: An Unsupervised Heterogeneous GNN Approach. WCNC 2025
Yi-Chien Lin*, Gangda Deng*, Viktor Prasanna
A Unified CPU-GPU Protocol for GNN Training.
Code
CF 2024
Gangda Deng*, Ömer Faruk Akgül*, Hongkuan Zhou, Hanqing Zeng, Yinglong Xia, Jianbo Li, Viktor Prasanna An Efficient Distributed Graph Engine for Deep Learning on Graphs. Code Slides SC Workshop 2023
Chenhuan Yu, Gangda Deng, Ning Gui PairGNNs: enabling graph neural networks with pair-based view Neural Computing and Applications 2023