Linghao (Lynn) JinΒ
Linghao (Lynn) JinΒ
I am a third-year PhD student in Computer Science at University of Southern California, advised by Professor Xuezhe Ma. Previously, I got my MS degree from Johns Hopkins University, advised by Benjamin Van Durme and Jerry L. Prince.
Currently I mostly engage in LLM pre-training, mid-training, Multimodal generation and Long-context modeling. Β I'm also deeply interested in continual learning with sparse architecture, complex reasoning in foundation models.Β
π¬ linghaoj [AT] usc [DOT] edu
[Feb 10th, 2025] Our paper "Progressive Compositionality in Text-to-Image Generative Models" got spotlight in ICLR 2025
[Sep 20th, 2024] Our paper "Light-weight fine-tuning for defending adversarial noise in medical VLMs" is accepted by EMNLP 2024
[Oct 7th, 2023] My first NLP paper is accepted by EMNLP 2023
K2-V2: A 360-Open, Reasoning-Enhanced LLM π
K2 Team, 2025
Progressive Compositionality in Text-to-Image Generative Models π
Xu Han, Linghao Jin, Xiaofeng Liu, Paul Pu Liang
ICLR 2025 (spotlight)
Exploring Opportunities to Support Novice Visual Artists' Inspiration and Ideation with Generative AI π
Cindy Peng, Alice Qian, Linghao Jin, Jieneng Chen, Evans Xu Han, Paul Pu Liang, Hong Shen, Haiyi Zhu, Jane Hsieh
Arxiv 2025
Long-context modeling
Gecko: An Efficient Neural Architecture Inherently Processing Sequences with Arbitrary Lengths π
Xuezhe Ma*, Shicheng Wen*, Linghao Jin*, Bilge Acun*, Ruihang Lai*, Bohan Hou, Will Lin, Hao Zhang, Songlin Yang, Ryan Lee, Mengxi Wu, Jonathan May, Luke Zettlemoyer, Carole-Jean Wu
Arxiv Preprint 2026
Context-aware Neural Machine Translation
Towards Chapter-to-Chapter Context-Aware Literary Translation via Large Language Models π
Linghao Jin, Li An, Xuezhe Ma
Arxiv Preprint 2024
Challenges in Context-Aware Neural Machine Translation π
Linghao Jin*, Jacqueline He*, Jonathan May, Xuezhe Ma
EMNLP 2023
MAX-ISI System at WMT23 Discourse-Level Literary Translation Task π
An Li*, Linghao Jin*, Xuezhe Ma
WMT 2023
Medical AI
Light-weight Fine-tuning Method for Defending Adversarial Noise in Pre-trained Medical Vision-Language Models π
Xu Han, Linghao Jin, Xuezhe Ma, Xiaofeng Liu
EMNLP (Findings) 2024
Effects of Defacing Whole Head MRI on Neuroanalysis π
Chenyu Gao, Linghao Jin, Jerry L Prince, Aaron Carass
SPIE 2022
Pattern Recognition
Mutual Information Regularized Identity-aware Facial Expression Recognition in Compressed Video π
Xiaofeng Liu, Linghao Jin, Xu Han, Jane You
Pattern Recognition 2021
Domain Generalization under Conditional and Label Shifts via Variational Bayesian Inference π
Xiaofeng Liu, Bo Hu, Linghao Jin, Xu Han, Fangxu Xing, Jinsong Ouyang, Jun Lu, Georges EL Fakhri, Jonghye Woo
IJCAI 2021
Rethinking the Invariant Feature Learning: Variational Bayesian Inference for Domain Generalization π
Xiaofeng Liu, B Hu, Linghao Jin, Xu Han, Fangxu Xing, Jinsong Ouyang, Jun Lu, Georges El Fakhri, Jonghye Woo
IJCAI 2021
Identity-aware Facial Expression Recognition in Compressed Video π
Xiaofeng Liu*, Linghao Jin*, Xu Han, Jun Lu, Jane You, Lingsheng Kong
ICPR 2020
[June 2025 - Aug 2025] Institute of Foundation Models (IFM)
Mentor:Β Hector Liu
Topics: LLM mid-training
[June 2024 - Sep 2024] Mohamed bin Zayed University of Artificial Intelligence: MBZUAI
Mentor:Β Hector Liu
Topics: Mixture of expert (MOE) model pre-training
2022 Fall: CSCI104 Data Structures and Object Oriented Design
2023 Spring: CSCI544 Applied Natural Language Processing
2023 Fall: CSCI 585 Database System
2024 Spring: CSCI544 Applied Natural Language Processing
[Jan, 24, 2024] "Challenges in Context-aware NMT"Β
@ Microsoft Reading group
[Nov, 2023] Poster PresentationΒ
@ EMNLP Singapore, 2023