cv
Basics
| Name | Yifan Yu |
| Date of Birth | November 30, 2001 |
| Languages | English, Chinese |
| Phone number | 470-845-7356 |
Education
-
2024 - Present PhD
University of Illinois Urbana-Champaign - Computer Science, Siebel School of Computing and Data Science
-
2020 - 2024 Bachelor of Science
Georgia Institute of Technology - Computer Science
- Threads
- Intelligence & Info Internetworks
Experience
-
2026.8 - Present Research Intern (Part-time)
Together AI - Working on efficient reinforcement learning for agentic diffusion language models and on combining looped Transformers with diffusion language modeling
-
2024 - Present Research Assistant
University of Illinois Urbana-Champaign - Working on efficient LLM serving, post-training, and multi-agent systems, advised by Prof. Fan Lai
-
2026.5 - 2026.8 Research Intern
Together AI - Led ORL (on-policy residual learning), co-designing the algorithm, training system, and inference engine for diffusion language models
- Single-request decoding 1.74x faster than speculative decoding (DFlash) at matched accuracy; up to 2,788 tokens/s on one RTX 5090 and 7,708 tokens/s on 8x B200
-
2026.1 - 2026.5 Research Intern (Part-time)
Together AI - Worked on introspective diffusion language models (I-DLM)
-
2025.5 - 2025.8 Applied Scientist Intern
Amazon - Worked on automatic error attribution for multi-agent systems
-
2024.5 - 2024.7 Student Researcher
Systems Research@Google - Worked on efficient LLM serving systems
-
2021.10 - 2024.5 Research Assistant
Georgia Institute of Technology - Increasing efficiency of LLMs
- Model compression with low rank approximation and pruning methods
- Quantization
- Increasing efficiency of LLMs
-
2022.5 - 2022.7 Data Scientist Intern
The Home Depot - Worked on effective retrieval with LLMs
Honors and Awards
-
2026 - COLM Oral Presentation (24 out of 2,900+ submissions)
-
2024 - ICLR Oral Presentation (86 out of 7,300+ submissions)
- Presidential Undergraduate Research Award
-
2020-2023 - Faculty Honor
Service
- Reviewer: NeurIPS 2025, ICLR 2026, ICML 2026, COLM 2026, NeurIPS 2026
Research Interests
-
Efficient machine learning, particularly efficient training and inference
- Diffusion language models
- Multi-agent systems
- Reinforcement learning
- Large language models
- Inference systems
Other Interests
- Guitar, Cooking, Work out, board games