Wafik Aboualim
I am a Master's thesis student at the University of Pittsburgh, where I mostly work on systems and reinforcement learning.
I am currently interested in efficient and sustainable training of foundation models across the training stack, including pre- and post-training. More recently, I have also been getting into embodied/physical AI.
Email: wat29 [at] pitt [dot] edu.
Research Interests
- Efficient training of foundation models
- Algorithms with predictions
- Optimization and resource allocation
- Quantum optimization
Papers
- Carbon-Aware Dynamic Parallelism for Distributed Training
ACM Energy Informatics Review, HOTCARBON 2026 proceedings. Volume 6, Issue 2, July 2026.
- Hybrid Combinatorial Multi-Armed Bandits for Discrete Resource Allocation Under Fairness Constraints
preprint. 2026. PDF
Recent Posts
News
- Paper accepted to HotCarbon 26'Our paper Carbon-Aware Dynamic Parallelism for Distributed Training was accepted to HotCarbon 26' and will appear in the ACM Energy Informatics Review proceedings.
- Joined the ReDSCAIP Lab at PittStarted working with Prof. Micheal Colaresi at the ReDSCAIP Lab on forecasting political violence.
- Started Thesis Master at PittBeginning my thesis Master program at the University of Pittsburgh.
- Selected for Qiskit Advocate Program 2.0Selected as part of the Qiskit Advocate Program 2.0, a global initiative recognizing individuals who actively contribute to the open-source Qiskit ecosystem and the broader quantum computing community.