- Human–AI Cognitive Alignment
- Trustworthy LLMs & Evaluation
- Expert-Aligned Conversational AI
- Eye-tracking & Physiological Sensing
- Cognitive Psychology of Trust
- Human-Centered AI Interfaces
PhD in NLP-HCI-Psychology
University of Amsterdam, the Netherlands
MSc in Optimization and AI (NLP)
Heidelberg University, Germany
BSc in Computational Mathematics
XiDian University, China
Does AI Think Like Humans?
Do Humans and AI share "cognitive" patterns and can biases pass between them?
Humans and LLMs lean on the same hidden cues — who wrote a text, not just what it says. That shared habit is how bias travels between the two sides.
How to Align AI with Human Signals?
Which human signals should enter a model — and which should be kept out?
Human data holds both expertise and bias. I align AI to keep the expert strategy, filter the bias, and adapt to the user.
Adaptive AI for Cognitive Augmentation
Can AI sense your trust and effort in real time — and adapt to truly support you?
AI senses human cognitive states from eye and body signals, then builds a closed-loop adaptive system for cognitive augmentation.
Together they form one bidirectional loop — understand the shared patterns, align AI with the right human signals, and augment the human in return — toward collaboration you can trust when it matters most: calibrated trust, appropriate reliance, and sharper decisions.
Two full papers were accepted to ACL 2026 (One Main, One Findings) in San Diego, USA.
One full papers were accepted to ACM UMAP 2026 in Gothenburg, Sweden.
A full paper and a poster were accepted to CHI 2026 in Barcelona.
New work was accepted by IJHCS and CSCW 2026.
A VR and AI project was selected for first-round incubation funding from the Wellcome Trust AI Accelerator.
A full paper and an LBW contribution were accepted to CHI 2025 in Yokohama.
See All
Ongoing publications and presentations across COLING, CSCW, ICMI, and leading HCI venues.