Binwen Liu

AI PhD Student • Xi'an Jiaotong University

Profile Photo

Hi! I'm a first-year direct-entry PhD student in Artificial Intelligence @ Xi'an Jiaotong University, advised by prof. Badong Chen.

Research

My research investigates structural gaps between human cognition and current AI systems—gaps that task-level benchmarks tend to obscure. I combine a critical lens, using cognitive science's problem framing to diagnose what is missing, with a constructive one, translating cognitive architecture principles into testable computational proposals.

Rather than committing to a single AI substrate, I examine how specific cognitive capacities—such as executive function, working memory, and metacognitive monitoring—manifest and fail differently across language models and action-taking agents, allowing cognitive claims to be triangulated across multiple empirical contexts.

I'm currently working on hierarchical representation learning based on taxonomical foundation for my Bachelor's Thesis:)

News

  • [1/2026]: Attend the kick-off meeting @ Fudan University as a participating member for the Research on a Brain-Inspired Multi-Agent Social Interaction System Driven by Affective Cognition and Moral Reasoning, approved under the Brain Science and Brain-Like Research National Science and Technology Major Project (2025-2030).
  • [10/2025]: Admitted to National Key Laboratory of Human-Machine Hybrid Augmented Intelligence & Institute of Artificial Intelligence and Robotics (IAIR) through Direct-Entry PhD program.

Selected Publications

Awards

2024
13th “China Software Cup” College Student Software Design Competition – Third Prize
2024
American Mathematical Contest in Modeling (MCM) – H Award
2023
"FLTRP●ETIC Cup" National College Foreign Language Contest – Silver Award (English Writing)

Services

Reviewer

[Conference 1 YYYY], [Conference 2 YYYY], [Journal 1 YYYY], [Journal 2 YYYY]

Teaching Assistant

[Course Code 1], [Course Code 2]

Volunteer

[Role 1], [Role 2], [Role 3]

Miscellaneous

Once inspired by Douglas Hofstadter & Josh Tenenbaum