CV
Curriculum vitae
Contact Information
| Name | Jakub Grzywaczewski |
| Professional Title | Research Software Engineer & Data Scientist |
| jakubzgrzywaczewski@gmail.com | |
| Phone | +48 790 200 186 |
| Location | Warsaw, Poland |
Professional Summary
Research Software Engineer at the Centre for Credible AI (CCAI) and MSc student in Mathematics and Information Science at Warsaw University of Technology, where I also graduated with honors in Data Science engineering. I focus on explainability of deep learning models using generative models, including diffusion models. Presented a scientific paper at ICLR 2025 and getting ready to present another one at ICML 2026. Former Python Developer at Samsung Research, where I worked on NLP projects and internal tooling.
Experience
-
10.2025 - present Warsaw, Poland
Research Software Engineer
Centre for Credible AI (CCAI)
Working on projects pushing the frontier of explainable AI, including practical application of generative AI for model validation and vision model depoisoning.
- Explained decision-making process of deep learning models using generative and diffusion models
- Co-authored a scientific paper which was accepted for ICML 2026
-
6.2024 - 10.2025 Warsaw, Poland
Research Software Engineer
MI2
Worked on explainability of deep learning models using generative and diffusion models.
- Co-authored and presented a scientific paper at ICLR 2025
-
8.2022 - 6.2024 Warsaw, Poland
Python Developer
Samsung Research
Worked on NLP research projects and developed internal CLI and web tools for data processing.
- Developed and maintained NLP pipelines from both research and implementation perspectives
- Built internal CLI and web tools for data processing
- Completed time management training and applied PRINCE2 methodology basics
Education
Publications
2026
-
Your CLIP has 164 dimensions of noise: Exploring the embeddings covariance eigenspectrum of contrastively pretrained vision-language transformersarXiv preprint arXiv:2605.14893, 2026 -
Auditing Sybil: Explaining Deep Lung Cancer Risk Prediction Through Generative Interventional AttributionsIn The Forty-Third International Conference on Machine Learning, 2026
2024
-
Rethinking visual counterfactual explanations through region constraintIn The Thirteenth International Conference on Learning Representations, 2024
Skills
Data Science: Python, NumPy, Pandas, Jupyter, Scikit-learn, PyTorch (Fabric), Transformers, SHAP, vLLM, Ray, ONNX, R, Tidyverse
Deep Learning & AI: Neural Networks, Generative Models, Diffusion Models, Explainable AI, Computer Vision, NLP
Databases: PostgreSQL, MySQL, MS SQL Server, Redis, Hadoop
Compute & Cloud: SLURM, Kubeflow, GCP, Kubernetes
Web & Mobile: JavaScript/TypeScript, Vue, Svelte, Dart (Flutter), Flask, FastAPI, Axum (Rust)
DevOps & Tooling: Git, Linux, Vim, Jenkins, GitHub Actions, Rust, Java, TeX
Languages
Polish : Native speaker
English : C1 Proficient
Interests
Generative Models for Explainable AI: Generative AI, Diffusion Models
Explainable AI: Model Validation, Interpretability, Depoisoning
Mathematics: Probability, Statistics, Linear Algebra
Hobbies: Ping-pong, Music, Cooking