Bio

I am a second-year master’s student in the Department of Computer Science and Engineering at the University of California San Diego. Prior to this, I studied at Xiamen Number One High School of Fujian, and obtained my bachelor’s degree in Artificial Intelligence from the School of the Gifted Young (SGY) at the University of Science and Technology of China (USTC).

I am interested in AI for science, and I am currently focused on neuroscience, medical imaging, and molecule interactions. My cross-disciplinary curiosity motivates me to solve real-world problems and advance scientific discovery using different machine learning tools.

Research

Measuring the Measures: Discriminative Capacity of Representational Similarity Metrics Across Model Families

Jialin Wu, Shreya Saha, Yiqing Bo, Meenakshi Khosla

Systematically evaluates the discriminative capacity of various representational similarity metrics across model families, revealing that geometry-preserving metrics yield stronger separation ability.

NeurIPS 2025 Workshop for UniReps• 3rd Edition of the Workshop on Unifying Representations in Neural Models

A Data-driven Typology of Vision Models from Integrated Representational Metrics

Jialin Wu, Shreya Saha, Yiqing Bo, Meenakshi Khosla

A biology-inspired framework that integrates multiple representational similarity metrics to create a principled typology of vision models.

arXiv preprint 2509.21628 • 2025 • Under Review

Integrated representational signatures strengthen specificity in brains and models

Jialin Wu, Shreya Saha, Yiqing Bo, Meenakshi Khosla

Demonstrates that integrating multiple representational similarity metrics through similarity network fusion yields substantially sharper discrimination of brain regions, revealing hierarchical organization that aligns with established anatomical and functional hierarchies.

arXiv preprint 2510.20847 • 2025 • Under Review

Saliency Guided Longitudinal Medical Visual Question Answering

Jialin Wu, Xiaofeng Liu

A saliency map guided encoder-decoder framework for longitudinal medical visual question answering to compare paired chest X-rays across time points, enforcing spatially consistent attention on corresponding anatomy for interpretable clinical change detection through having one step for saliency generation and the other step for answer generation.

NeurIPS 2025 Workshop for Imageomics • Discovering Biological Knowledge from Images Using AI

DINO-Inspired Attention-Guided Longitudinal Medical Visual Question Answering System

Jialin Wu, Xiaofeng Liu

An attention-guided encoder-decoder framework for longitudinal medical visual question answering that combines DINO-based adaptive mask generation and auxiliary DINO-inspired learning objectives to achieve strong performance on the Medical-Diff-VQA with intrinsic interpretability.

Under Review

Benchmark on Drug Target Interaction Modeling from a Structure Perspective

Xinnan Zhang*, Jialin Wu*, Junyi Xie, Tianlong Chen, Kaixiong Zhou

A comprehensive benchmark for drug-target interaction (DTI) modeling from a structure perspective, evaluating various computational methods and providing 9 key insights in DTI task.

arXiv preprint 2407.04055 • 2024 • Under Review

Other Papers

Possibilities and Implications of the Multi-AI Competition

Jialin Wu

An exploration of the potential scenarios and implications when multiple AI systems compete with each other, analyzing the impact on AI safety and alignment.

arXiv preprint 2209.00509 • 2022

Nitrogen Association

Jialin Wu, Qizhi Yao

A Chinese science popularization article discussing nitrogen compounds.

University Chemistry.

Skills

C

Programming Language

Python

Programming Language

Java

Programming Language

MATLAB

Scientific Computing

Docker

Containerization

Premiere

Video Editing

Honors & Awards

Services & Leadership

More About Me