01
Self-Evolving AI
Continual self-improvement, model adaptation, model merging, and autonomous learning under changing data and tasks.
Khalifa University · Abu Dhabi
Self-Evolving &
Adaptive Intelligence
We build AI systems that learn to adapt, reason, improve, and remain reliable in changing environments.
Self-Evolving AIMultimodal IntelligenceReasoning & AgentsTrustworthy AI
What we work on
Research
Our research spans learning, reasoning, multimodal perception, efficiency, and reliability — with a common goal of enabling AI systems to adapt and improve over time.
02
Multimodal Intelligence
Robust learning across vision, language, audio, and medical modalities, including incomplete and uncertain observations.
03
Reasoning & Agents
Reinforcement learning, language-model reasoning, planning, and agentic systems that learn from interaction and feedback.
04
Efficient Foundation Models
Pruning, quantization, prompting, and efficient adaptation for large vision-language and generative models.
05
Trustworthy AI
Robustness, privacy, security, uncertainty estimation, and reliable deployment of learning systems.
06
AI for Healthcare
Multimodal medical AI for clinically meaningful perception, diagnosis, and decision support.
Latest
News
BMVC 2026 — Rethinking Weight-Averaged Model-merging accepted.
CVPR 2026 — TransPrune accepted for efficient large vision-language models.
CVPR 2026 — Polyphony accepted for dual-hand action segmentation.
ICASSP 2026 Oral — Work on reducing prompt sensitivity in vision-language models accepted.
LLM Reasoning — Reinforcement-learning-based language model reasoning work released.
People
Members
SAIL brings together researchers working on adaptive, multimodal, efficient, and trustworthy intelligence.
Director
Hu Wang
Lab Director · Assistant Professor
Khalifa University
Multimodal learning · Reinforcement learning · Medical AI · AI security
Students & Alumni
HW
Hanwen Wang
PhD Student
University of Adelaide
Co-supervising with Dr. Tim Chen
MA
Maryam Arjemandi
Master's Student
MBZUAI
Co-supervised with Prof. Mohammad Yaqub
DH
Dongli He
Master's Student
MBZUAI
Co-supervised with Prof. Mohammad Yaqub
MP
Maxim Popov
Master's Student
MBZUAI
Co-supervised with Prof. Mohammad Yaqub
YZ
Yuan Zhang
PhD Student
University of Adelaide
Co-supervised with Prof. Gustavo Carneiro
RW
Renjie Wu
Master's Student
University of Adelaide
Co-supervised with Dr. Tim Chen
ZW
Zihan Wang
Master's Student
University of Adelaide
Co-supervised with Dr. Jason Xue
DB
David Butler
Master's Student
University of Adelaide
Co-supervised with Prof. Gustavo Carneiro
Selected work
Publications
TransPrune: Token Transition Pruning for Efficient Large Vision-Language Model
A. Li, Y. Duan, J. Zhang, C. Ma, Y. Xie, G. Carneiro, M. Yaqub, H. Wang
Polyphony: Diffusion-based Dual-Hand Action Segmentation
H. Zheng, H. Wang, T. Zheng, P. Bhattarai, T. Alhanai
Rethinking Weight-Averaged Model-merging
H. Wang, C. Ma, I. Almakky, I. Reid, G. Carneiro, M. Yaqub
Multi-modal Learning with Missing Modality via Shared-Specific Feature Modelling
H. Wang, Y. Chen, C. Ma, J. Avery, L. Hull, G. Carneiro
Uncertainty-aware Multi-modal Learning via Cross-modal Random Network Prediction
H. Wang, J. Zhang, Y. Chen, C. Ma, J. Avery, L. Hull, G. Carneiro
Join the lab
Build the next generation of adaptive intelligence.
We welcome highly motivated students and collaborators interested in AI, including prospective Master's and PhD students at Khalifa University.
Contact Dr. Hu Wang