Controllable generation
Example-driven terrain synthesis and graph-based sketching, with user control over terrain geometry, semantic coverage, and materials.
Yunyu Liu
Ph.D. Student · Computer Science · Purdue University
I am a Ph.D. student at Purdue University, advised by Prof. Bedrich Benes. My research focuses on generative modeling, controllable terrain authoring, and graph representation learning.
At Meta BizAI, I completed machine-learning internships in 2024 and 2025, working on multilingual generation, hallucination control, reward modeling, and reinforcement learning.
01 / RESEARCH
Research
Reliable language models
Multilingual generation and hallucination control, with reward modeling and reinforcement learning for advertising applications.
Learning on graphs
Temporal networks, higher-order interactions, hypergraph diffusion, and representations that capture relational structure.
Work in Progress
A Unified Framework for Multi-Turn Vegetation Authoring
Work in progress
An interactive framework for editing vegetation layouts over multiple turns. It combines reference-distribution controls with direct controls over effects such as visibility and shading, while retaining earlier goals and placement constraints as new edits are introduced. Users can review results and explicitly revise preferences to guide subsequent edits.
Land-Cover Authoring
Work in progress
I study controllable generation and editing of semantic land-cover layouts for digital landscapes. This research explores how terrain elevation and high-level user preferences can guide plausible coverage patterns and regional edits.
02 / EXPERIENCE
Experience
Meta · BizAI
Software Engineer Intern, Machine Learning
Developed hallucination-control components for multilingual advertising text and evaluated generated content for safety and reliability.
Meta · BizAI
Software Engineer Intern, Machine Learning
Designed reward-modeling components and applied PPO-style reinforcement learning to optimize advertising text under product constraints.
Education
Purdue University
Ph.D. in Computer Science · Advisor: Bedrich Benes
CGV Lab: terrain generation and sketch-based authoring. Earlier research at GCoM Lab: temporal networks and higher-order interaction prediction.
Northeastern University
M.S. in Computer Engineering
Multi-view and semi-supervised learning in Prof. Yun Raymond Fu's Synergetic Media Learning Lab.
Shanghai Jiao Tong University
B.Eng. in Electrical Engineering · Minor in Finance
LatentMimic: Terrain-Adaptive Locomotion via Latent Space Imitation
Zhiquan Wang, Yunyu Liu, Dipam Patel, Ayush Kumar, Aniket Bera, Bedrich Benes
MIGSingle-Shot Example Terrain Sketching by Graph Neural Networks
Yunyu Liu, Bedrich Benes
Computer Graphics Forum, 44(1), e15281Human Motion Segmentation via Velocity-Sensitive Dual-Side Auto-Encoder
Yue Bai, Lichen Wang, Yunyu Liu, Yu Yin, Hang Di, Yun Fu
IEEE Transactions on Image Processing, 32:524–536Equivariant Hypergraph Diffusion Neural Operators
Peihao Wang, Shenghao Yang, Yunyu Liu, Zhangyang Wang, Pan Li
International Conference on Learning Representations (ICLR)Neural Predicting Higher-Order Patterns in Temporal Networks
Yunyu Liu, Jianzhu Ma, Pan Li
ACM Web Conference (WWW), pp. 1340–1351Semi-Supervised Dual Relation Learning for Multi-Label Classification
Lichen Wang, Yunyu Liu, Hang Di, Can Qin, Gan Sun, Yun Fu
IEEE Transactions on Image Processing, 30:9125–9135Aspect-Based Sentiment Classification via Reinforcement Learning
Lichen Wang, Bo Zong, Yunyu Liu, Can Qin, Wei Cheng, Wenchao Yu, Xuchao Zhang, Haifeng Chen, Yun Fu
IEEE International Conference on Data Mining (ICDM), pp. 1391–1396Inductive Representation Learning in Temporal Networks via Causal Anonymous Walks
Yanbang Wang, Yen-Yu Chang, Yunyu Liu, Jure Leskovec, Pan Li
International Conference on Learning Representations (ICLR)GENPass: A Multi-Source Deep Learning Model for Password Guessing
Zhiyang Xia, Ping Yi, Yunyu Liu, Bo Jiang, Wei Wang, Ting Zhu
IEEE Transactions on Multimedia, 22(5):1323–1332Dual-Side Auto-Encoder for High-Dimensional Time Series Segmentation
Yue Bai, Lichen Wang, Yunyu Liu, Yu Yin, Yun Fu
IEEE International Conference on Data Mining (ICDM), pp. 918–923Dual Relation Semi-Supervised Multi-Label Learning
Lichen Wang, Yunyu Liu, Can Qin, Gan Sun, Yun Fu
AAAI Conference on Artificial Intelligence (AAAI), pp. 6227–6234Generative View-Correlation Adaptation for Semi-Supervised Multi-View Learning
Yunyu Liu, Lichen Wang, Yue Bai, Can Qin, Zhengming Ding, Yun Fu
European Conference on Computer Vision (ECCV), pp. 318–334Generative Multi-View Human Action Recognition
Lichen Wang, Zhengming Ding, Zhiqiang Tao, Yunyu Liu, Yun Fu
IEEE/CVF International Conference on Computer Vision (ICCV), pp. 6211–6220GENPass: A General Deep Learning Model for Password Guessing with PCFG Rules and Adversarial Generation
Yunyu Liu, Zhiyang Xia, Ping Yi, Yao Yao, Tiantian Xie, Wei Wang, Ting Zhu
IEEE International Conference on Communications (ICC), pp. 1–6Password Guess and Analyze Based on Recurrent Neural Network
Zhiyang Xia, Yunyu Liu, Ping Yi
Conference on Vulnerability Analysis and Risk Assessment (VARA)