Hello, I’m Junjie Oscar Yin#
My research is in NLP and ML. Specifically, I am interested in understanding the relation between training data and language model’s abilities beyond simple scaling, and, through that, do good science on AI systems.
Recently, my friends and I started WAI, where we work on open agentic research.
Bio#
I am a first-year PhD student at UW NLP, advised by Hannaneh Hajishirzi. Previously, I was mentored by Sasha Rush at Cornell.
In my undergraduate years at Johns Hopkins, I read EE, CS, and Philosophy. I was grateful to do research under Alan Yuille and with the support of the Westgate Scholarship. I also spent a year at Oxford, where I studied under Michael Wooldridge.
Publications#
My list of publications is available here:
- TMax: A Simple Receipe for Terminal Agents
Hamish Ivison*, Junjie Oscar Yin*, Rulin Shao, Teng Xiao, Nathan Lambert, Hannaneh Hajishirzi.1
Preprint
[Paper] [Codebase] [Blog Post]
Learning to Detect Language Model Training Data via Active Reconstruction.
Junjie Oscar Yin, John X. Morris, Vitaly Shmatikov, Sewon Min, Hannaneh Hajishirzi.
Preprint
[Paper] [Codebase]Approximating Language Model Training Data from Weights.
John X. Morris, Junjie Oscar Yin, Woojeong Kim, Vitaly Shmatikov, Alexander M. Rush.
Conference on Language Modeling (COLM), 2025.
[Paper] [Codebase]Compute-Constrained Data Selection.
Junjie Oscar Yin, Alexander M. Rush.
International Conference on Learning Representations (ICLR), 2025.
[Paper] [Codebase] [Talk]ModuLoRA: Finetuning 2-Bit LLMs on Consumer GPUs by Integrating with Modular Quantizers.
Junjie Oscar Yin, Jiahao Dong, Yingheng Wang, Christopher De Sa, Volodymyr Kuleshov.
Transactions on Machine Learning Research (TMLR), 2024. (Featured Paper ; Presented at ICLR 2024)
[Paper] [Codebase] [Blog Post]
Recent Updates#
Below you will find the most recent updates
2025 Started my PhD @UWNLP
2024 Computing Research Association’s (CRA) Outstanding Undergraduate Researcher Award (URA). Dec 2023.
*Co-first authors. ↩︎