Tian Yu Liu

I am a PhD student at UCLA Vision Lab, advised by Prof. Stefano Soatto. My research interests lie in computer vision and machine learning. I also obtained my Bachelor of Science (Summa Cum Laude) from UCLA with a major in Computer Science and minor in Mathematics, where I worked with Prof. Baharan Mirzasoleiman and Prof. Alex Wong.

Google Scholar  /  Github  /  LinkedIn  /  tianyu (at) cs (dot) ucla (dot) edu

Website last updated: Mar 2025

profile photo
Selected Publications (see full list here)
PICASO: Permutation-Invariant Context Composition with State Space Models
Tian Yu Liu, Alessandro Achille, Matthew Trager, Aditya Golatkar, Luca Zancato, Stefano Soatto
International Conference on Learning Representations (ICLR), 2025
[Paper]
Conjuring Semantic Similarity
Tian Yu Liu and Stefano Soatto
International Conference on Learning Representations (ICLR) Workshop on Representation Alignment, 2025
[Paper]
Meanings and Feelings of Large Language Models: Observability of Latent States in Generative AI
Tian Yu Liu, Stefano Soatto, Matteo Marchi, Pratik Chaudhari, Paulo Tabuada
International Conference on Learning Representations (ICLR) Workshop on Bidirectional Human-AI Alignment, 2025
[Paper]
Interpretable Measures of Conceptual Similarity by Complexity-Constrained Descriptive Auto-Encoding
Alessandro Achille, Greg Ver Steeg, Tian Yu Liu, Matthew Trager, Carson Klingenberg, Stefano Soatto
Conference on Computer Vision and Pattern Recognition (CVPR), 2024.
[Paper]
Meaning Representations from Trajectories in Autoregressive Models
Tian Yu Liu, Matthew Trager, Alessandro Achille, Pramuditha Perera, Luca Zancato, Stefano Soatto.
International Conference on Learning Representations (ICLR), 2024
[Paper][Code]
Tangent Transformers for Composition, Privacy and Removal
Tian Yu Liu, Aditya Golatkar, Stefano Soatto
International Conference on Learning Representations (ICLR), 2024
[Paper][Code]
Tangent Model Composition for Ensembling and Continual Fine-tuning
Tian Yu Liu and Stefano Soatto
International Conference on Computer Vision (ICCV), 2023.
[Paper] [Code] [Video]
Friendly Noise against Adversarial Noise: A Powerful Defense against Data Poisoning Attacks
Tian Yu Liu, Yu Yang, Baharan Mirzasoleiman
Advances in Neural Information Processing Systems (NeurIPS), 2022.
[Paper] [Code]
Data-Efficient Augmentation for Training Neural Networks
Tian Yu Liu and Baharan Mirzasoleiman
Advances in Neural Information Processing Systems (NeurIPS), 2022.
[Paper] [Code]
Monitored Distillation for Positive Congruent Depth Completion
Tian Yu Liu, Parth Agrawal, Allison Chen, Byung-Woo Hong, Alex Wong
European Conference on Computer Vision (ECCV), 2022.
[Paper] [Code]

I stole this nice template from here.