Adam Klukowski
Adam Klukowski is a Polish mathematician and machine-learning researcher. He won gold at the International Mathematical Olympiad 2015 and silver in 2016 for Poland (plus IChO silver 2016), then read mathematics at Cambridge (BA, MASt). He worked as a machine-learning researcher at Huawei's UK R&D / Noah's Ark Lab (2020-2022) on deep-learning theory, briefly as an ML engineer at CHARM Therapeutics (2022), and completed a PhD in pure mathematics (geometric topology) at the University of Oxford. As of September 2026 he is a postdoctoral fellow at the Max Planck Institute for Mathematics in Bonn. His only quantitative-finance role was a 2019 summer Quantitative Research Internship at Citadel Securities in London.
- Quant employment verification: CONFIRMED. Actual quant-finance internship: Quantitative Research Intern at Citadel Securities, London, 2019 (LinkedIn experience section), in the gap between his Cambridge BA and MASt; Citadel Securities is a genuine quantitative market-maker (systematic electronic trading), so this is real quant employment, not a student program. His other roles are NOT quant-finance: Samsung engineering intern 2017, Cambridge chemistry summer researcher 2018, Huawei ML researcher 2020-2022, CHARM Therapeutics ML engineer 2022. No other trading/hedge-fund/asset-manager/energy/crypto-trading role is evidenced; identity confirmed by the co-located Olympiad, Cambridge/Oxford and Huawei records.Jul 2019
- Now postdoctoral fellow at the Max Planck Institute for Mathematics, Bonn (2026-09-01 to 2028-08-31; topology/low-dimensional topology/mapping class groups) — a decisive move from industry back to a full academic research track.Sep 2026
- IMO record: represented Poland twice, gold 2015 (31 pts) and silver 2016 (23 pts) as a student of XIV LO im. Staszica, Warsaw — the identity anchor tying the profile to a nationally elite math record.Jul 2016
- LinkedIn still surfaces CHARM Therapeutics as his top experience while the MPIM listing and his 2022-2026 PhD timeline postdate it; the index is stale on his present role, so employment here is recorded from the MPIM page with that conflict flagged rather than asserted from the index.Sep 2026
- Cross-source read: three phases — competitor (IMO/IChO 2015-2016), industry ML (Samsung 2017, Cambridge 2018, Huawei 2020-2022, CHARM 2022), and pure maths (Oxford PhD 2022-2026, MPIM 2026) — with the sole quant detour a 2019 Citadel internship, so a quant-finance career is NOT the throughline despite the quant-coded lead.Sep 2026
- PhD in Mathematics at Oxford (2022-2026) on geometric topology; thesis/central paper 'Simple closed curves, non-kernel homology and Magnus embedding' (arXiv 2023; LMS 2025), a genuine research doctorate.Apr 2023
- Cambridge mathematics (BA 2016-2019, MASt 2019-2020) then a Citadel Securities quant internship in the 2019 summer gap — the linear elite-maths-to-quant-firm pipeline, before he veered to research instead.2020
- Supervised at Oxford by Martin Bridson with Dawid Kielak co-supervising (Kielak's page lists him as a student; a 2025 paper thanks 'his Ph.D. supervisor Martin Bridson for general guidance'), fixing his doctoral lineage in geometric group theory.2025
- Huawei Noah's Ark Lab deep-learning-theory papers ('Exact Count of Boundary Pieces of ReLU Classifiers'; 'Rate of Convergence of Polynomial Networks to Gaussian Processes', UAI 2022) show his 2020-2022 industry phase was genuine ML research, not engineering support — a math talent redirected into ML theory before returning to pure maths.Jun 2022
- Also took silver at the International Chemistry Olympiad 2016, a rare math-plus-chemistry double showing breadth beyond a single-subject specialist.Jul 2016
- Active seminar presence in his PhD years (KCL maths department 10 Jun 2025; Oxford Geometric Group Theory seminar on tangles in random covers) — talks that mark a researcher already circulating in the field.Jun 2025
- 'Tangle Free Permutations and the Putman-Wieland Property', IMRN 2024, is collaborative journal-placed work — evidence of a functioning research group around him, not a lone student.Oct 2024
Data-driven drug design.
Researching the theory of deep learning. See "Exact Count of Boundary Pieces of ReLU Classifiers: Towards the Proper Complexity Measure for Classification" and "Rate of Convergence
I did a research project about path integrals in the Department of Chemistry of University of Cambridge. Among the results was a criterion for agreement between quantum path integr
I was assessing and troubleshooting wireless hardware and software. It involved setting up and overseeing automated tests. Together with my colleague we found the reason of underpe
Competition record
Poland · IMO