Education and Research Visits ๐
๐๏ธ Oct 2021 โ Present ยท Warsaw, Poland ๐ต๐ฑ
PhD in Deep Learning at Warsaw University of Technology
I am currently pursuing a PhD in Deep Learning under the supervision of professor Tomasz Trzciลski. My research focuses on efficient machine learning methods, including adaptive computation, early exits, activation sparsity, mixture-of-experts architectures, continual learning, and knowledge distillation. The PhD program is co-sponsored by IDEAS NCBR, ELLIS Unit Warsaw.
๐๏ธ OctโNov 2023 ยท Rome, Italy ๐ฎ๐น
Visiting PhD Student at Sapienza University of Rome
Visited professor Simone Scardapane’s group in Rome to work on adaptive computation techniques in deep learning. The visit led to several joint publications with professor Scardapane, including “Universal Properties of Activation Sparsity in Modern Large Language Models” and “Exploiting Activation Sparsity with Dense to Dynamic-k Mixture-of-Experts Conversion” (NeurIPS 2024).
๐๏ธ 2022โ2023 ยท Krakรณw, Poland ๐ต๐ฑ
PhD Student Collaborator at Jagiellonian University
Collaborated on the Bio-inspired Artificial Neural Networks grant (BiONN) at Jagiellonian University. The project led to long-term collaborations and several publications with the GMUM group from Krakรณw.
๐๏ธ Feb 2019โSep 2020 ยท Warsaw, Poland ๐ต๐ฑ
Master’s (MEng) in Computer Science at Warsaw University of Technology
Thesis: “Exploitation of Deep Neural Network Architectures for Audio Enhancement”
Graduated with highest honors (5.0/5.0). The thesis investigated the use of deep learning techniques to enhance audio quality. The project is available in the GitHub repository, which includes implementations of autoencoders, SEGAN, and WaveNet for audio denoising and upsampling.
๐๏ธ Sep 2015โFeb 2019 ยท Warsaw, Poland ๐ต๐ฑ
Bachelor’s (BEng) in Industrial Informatics at Warsaw University of Technology
Thesis: “Keyword Spotting System Based on TensorFlow Library”
Graduated with highest honors (5.0/5.0). The thesis focused on developing a keyword spotting system for speech using the then-most popular deep learning framework, TensorFlow. As part of the project, I developed a demo that allowed users to control a simulated robot via voice commands.
Grants & Awards ๐
๐๏ธ 2025โ2026 ยท Poland ๐ต๐ฑ
Principal Investigator in a PRELUDIUM Grant, Polish National Science Center
“Continual Learning with Conditional Computation Networks”
Authored and led a project on adaptive computation for continual learning as Principal Investigator. PRELUDIUM grants are the most prestigious national grants for early-career scientists in Poland, and my proposal was ranked 2nd nationwide among machine learning projects in 2024.
๐๏ธ 2022โ2026 ยท Poland ๐ต๐ฑ
PhD Student in an OPUS Grant, Polish National Science Center
“Dynamic Neural Networks for Efficient Machine Learning”
Co-authored the grant proposal on the adaptive computation side and worked in the project as a PhD student.
๐๏ธ 2024 ยท Warsaw, Poland ๐ต๐ฑ
Scientific Award for PhD Students, Warsaw University of Technology
Received the annual award recognizing outstanding doctoral research.
Reviewing Experience ๐
๐๏ธ 2026 โ ICLR 2026 - Currently enrolled as a reviewer for the main track
๐๏ธ 2025 โ NeurIPS 2025 - Reviewed 5 papers for main track
๐๏ธ 2025 โ ICML 2025 - Reviewed 4 papers for the main track
๐๏ธ 2022 โ ICML Workshop on Dynamic Neural Networks - Reviewed several workshop submissions
๐๏ธ 2023, 2024, 2025 โ ML in PL Conference - Reviewed poster and talk submissions
๐๏ธ 2022, 2023 โ ML Summer School Krakรณw - Reviewed candidate applications
Teaching & Mentoring ๐
๐๏ธ May 2025 ยท Warsaw, Poland ๐ต๐ฑ
Guest Speaker at Warsaw University of Technology
Presented “Inference acceleration and adaptive computation in deep neural networks” for the [GOLEM student society].
๐๏ธ June 2024 ยท Rome, Italy ๐ฎ๐น
Teaching Assistant at Sapienza University of Rome
Served as teaching assistant for the [Deep Generative View on Continual Learning course], focusing on generative approaches to continual learning.
๐๏ธ Spring semesters of 2023, 2024, 2025 ยท Warsaw, Poland ๐ต๐ฑ
Teaching Assistant at Warsaw University of Technology
Served as TA for the Introduction to Artificial Intelligence course across three semesters, supervising student projects and conducting lab sessions with approximately 60 students per year. The course covered fundamental AI topics including optimization techniques, basic machine learning algorithms and neural networks.
๐๏ธ 2023โ2025 ยท Warsaw, Poland ๐ต๐ฑ
Student Supervision at Warsaw University of Technology
Supervised and co-supervised three student projects that resulted in published research and BSc/MSc theses: “ExpertSim: Fast Particle Detector Simulation Using Mixture-of-Generative-Experts” (ECAI 2025), “Exploring the Stability Gap in Continual Learning: The Role of the Classification Head” (WACV 2025), and “Exploiting Activation Sparsity with Dense to Dynamic-k Mixture-of-Experts Conversion” (NeurIPS 2024).