Scholar
Tianyu Chen
Google Scholar ID: 8Aum3V8AAAAJ
University of Texas at Austin
Citations & Impact
Citations
69
H-index
3
i10-index
2
Publications
14
Co-authors
11
list available
Publications
Resume (English only)
Academic Achievements
- 2025 preprint 'Denoising Score Distillation: From Noisy Diffusion Pretraining to One-Step High-Quality Generation' proposes training a one-step image generator using only noisy images, achieving FID scores comparable to clean-image-trained diffusion models; demonstrates distillation enhances generation quality beyond mere acceleration
- Paper 'Conditional diffusions for amortized neural posterior estimation' accepted by AISTATS 2025
- Two papers accepted by NeurIPS 2024: 'Diffusion Policies creating a Trust Region for Offline Reinforcement Learning' and 'Identifying General Mechanism Shifts in Linear Causal Representations'
- Published 'iSCAN: identifying causal mechanism shifts among nonlinear additive noise models' at NeurIPS 2023
- Published 'Model-based trajectory inference for single-cell rna sequencing using deep learning with a mixture prior' in PNAS
Background
- Second-year PhD student at the University of Texas at Austin, supervised by Professor Mingyuan Zhou
- Broad research interests in statistical machine learning, specifically:
- - Generative Models (e.g., diffusion models, distillation, VAEs, inverse problems)
- - Reinforcement Learning for LLMs (e.g., offline/online RL for training LLMs)
- - Causal Inference (e.g., graphical probabilistic models, causal representation learning)
- - Statistical Sampling (e.g., neural posterior sampling)
- - Bioinformatics (e.g., single-cell and multi-omic data integration)
- Open to Summer Internship 2025 opportunities
