PhD Student in Computer Science, UNSW SydneyVisiting Researcher, NExT++ Latent Knowledge Group, NUS
Mill Liao (Chih-Ting Liao) is a PhD student in Computer Science at the School of Computer Science and Engineering, University of New South Wales, and a Visiting Researcher with the NExT++ Latent Knowledge Group at the National University of Singapore. His research centers on mechanistic interpretability and latent knowledge in vision-language (VLM) and vision-language-action (VLA) models, with the broader goal of building trustworthy multimodal reasoning systems. He studies how models internally represent spatial and temporal structure, when their reasoning is genuinely grounded in visual evidence rather than driven by priors, and how these internal mechanisms can be measured and repaired. He is currently also a Research Intern at Crest Robotics.
Warning
Problem: The current name of your GitHub Pages repository ("Solution: Please consider renaming the repository to "
http://".
However, if the current repository name is intended, you can ignore this message by removing "{% include widgets/debug_repo_name.html %}" in index.html.
Action required
Problem: The current root path of this site is "baseurl ("_config.yml.
Solution: Please set the
baseurl in _config.yml to "Education
-
University of New South WalesSchool of Computer Science and Engineering (CSE)
PhD in Computer ScienceSep. 2025 - present -
University of SydneyMaster of Data Science (Distinction)Sep. 2020 - Jul. 2021
Honours & Awards
-
Engineering Domestic Research Excellence Scholarship (EDRES), Faculty of Engineering, University of New South Wales2025
-
Australian Government Research Training Program (RTP) Stipend and Faculty of Engineering and IT Top-Up, University of Technology Sydney2025
News
Selected Publications (view all )
SpaMEM: Benchmarking Dynamic Spatial Reasoning via Perception-Memory Integration in Embodied Environments
Chih-Ting Liao, X. Xiao, C. Meng, Z. Chen, Y. Qiao, W. Zhou, T. Wang, X. Zheng
European Conference on Computer Vision (ECCV) 2026
TODO: add a 1~2 sentence TLDR. A benchmark probing dynamic spatial reasoning of embodied vision-language models through perception-memory integration.
SpaMEM: Benchmarking Dynamic Spatial Reasoning via Perception-Memory Integration in Embodied Environments
Chih-Ting Liao, X. Xiao, C. Meng, Z. Chen, Y. Qiao, W. Zhou, T. Wang, X. Zheng
European Conference on Computer Vision (ECCV) 2026
TODO: add a 1~2 sentence TLDR. A benchmark probing dynamic spatial reasoning of embodied vision-language models through perception-memory integration.
Staying VIGILant: Mitigating Visual Laziness via Counterfactual Visual Alignment in MLLMs
X. Xiao, C. Liu, Chih-Ting Liao, Y. Zhang, Q. Lan, Y. Wei, L. Zhao, J. Wang, J. Gu
European Conference on Computer Vision (ECCV) 2026
TODO: add a 1~2 sentence TLDR of VIGILant here.
Staying VIGILant: Mitigating Visual Laziness via Counterfactual Visual Alignment in MLLMs
X. Xiao, C. Liu, Chih-Ting Liao, Y. Zhang, Q. Lan, Y. Wei, L. Zhao, J. Wang, J. Gu
European Conference on Computer Vision (ECCV) 2026
TODO: add a 1~2 sentence TLDR of VIGILant here.
Reasoning-VLA: A Fast and General Vision-Language-Action Reasoning Model for Autonomous Driving
D. Zhang, Z. Yuan, Z. Chen, Chih-Ting Liao, Y. Chen, F. Shen, Q. Zhou, Tat-Seng Chua
International Conference on Machine Learning (ICML) 2026
TODO: add a 1~2 sentence TLDR of Reasoning-VLA here.
Reasoning-VLA: A Fast and General Vision-Language-Action Reasoning Model for Autonomous Driving
D. Zhang, Z. Yuan, Z. Chen, Chih-Ting Liao, Y. Chen, F. Shen, Q. Zhou, Tat-Seng Chua
International Conference on Machine Learning (ICML) 2026
TODO: add a 1~2 sentence TLDR of Reasoning-VLA here.