CV
Tingting Du
Computer Science and Mathematics undergraduate at UW–Madison
Summary
Robot learning researcher working on trimanual manipulation, human–robot collaboration, Vision-Language-Action models, and LLM reasoning.
Education
- B.S. in Computer Science and Mathematics2027-05 (expected)University of Wisconsin–MadisonGPA: 3.86/4.00Courses: Autonomous Robotics, Operating Systems, Neural Networks
- Visiting Student in Computer Science2024-06University of California, BerkeleyGPA: 3.77/4.00Courses: Real Analysis, Data Structures, Algorithms, Artificial Intelligence
- Undergraduate studies in Linguistics2023-06Ningbo UniversityGPA: 3.80/4.00
Work Experience
- Trimanual Manipulation Research2026-08 - PresentRT² Lab, University of Wisconsin–MadisonAdvisor: Prof. Mike Hagenow
- Learning from demonstration for trimanual human–robot team manipulation.
- Wired a unified emergency-stop interlock tying all three arms to a single safety circuit.
- Vision-Language-Action and World Action Models2025-11 - 2026-08CASE Lab, University of MarylandAdvisor: Prof. Ang Li
- Developed a multi-layer shared projector bridging 2D and 3D representations in VLA models (NeurIPS 2026).
- Developed XGenAct, a geometry-enhanced world action model (under review).
- Deployed and evaluated both systems on real-world manipulation with a UFACTORY xArm.
- Student Modeling for Question Generation2024-07 - 2025-01DM² Lab, University of Notre DameAdvisor: Prof. Meng Jiang
- Developed a student modeling approach for question generation (ACL 2025).
- Situated Language Use and Convention Formation2024-02 - 2024-06Berkeley Artificial Intelligence ResearchAdvisor: Prof. Alane Suhr
- Analyzed ad-hoc convention formation in a large 3D-environment dialogue corpus.
Skills
Robotics
- ROS 2
- UFACTORY xArm
- Multi-arm workcell bring-up and safety integration
- Onshape CAD
- 3D printing
Machine Learning
- PyTorch
- Hugging Face Transformers
- vLLM
- PEFT/LoRA
- RLBench
Languages and Infrastructure
- Python
- C/C++
- Java
- Julia
- R
- JavaScript/TypeScript
- SQL
- Assembly
- LaTeX
- Linux
- Docker
- AWS
Publications
- XGenAct: Geometry-Enhanced World Action Models through Cross-Task Generation2026arXiv:2610.03516A geometry-enhanced world action model that learns perception and action streams through cross-task generation.
- ROCKET: Residual-Oriented Multi-Layer Alignment for Spatially-Aware Vision-Language-Action Models2026Advances in Neural Information Processing Systems (NeurIPS)A residual-oriented multi-layer alignment framework for injecting 3D spatial reasoning into VLA models.
- Vision-Language-Action in Robotics: A Survey of Datasets, Benchmarks, and Data Engines2026Transactions on Machine Learning Research (TMLR)A survey of the data infrastructure supporting Vision-Language-Action models in robotics.
- Agent KB: A Hierarchical Memory Framework for Cross-Domain Agentic Problem Solving2025ICML 2025 Workshop on Collaborative and Federated Agentic WorkflowsA hierarchical memory framework for cross-domain agentic problem solving.
- QG-SMS: Enhancing Test Item Analysis via Student Modeling and Simulation2025Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (ACL)A student modeling and simulation approach for question generation and test item analysis.
- Characterizing Language Use in a Collaborative Situated Game2025arXiv:2512.03381An analysis of situated language use and ad-hoc convention formation in a collaborative 3D environment.
Interests
- ResearchRobotics, Human–Robot Collaboration, Vision-Language-Action Models, LLM Reasoning