- Tapping into Rhythm Music & Science, 2026
- Motivation & Working Memory BSc thesis, RUG, 2021
- Consciousness in Animals BA thesis, RUG, 2021
Posts
- Nothing posted yet.
Content goes here
tbd
CV
Computational modelling of complex systems — building and validating statistical models of human and collective behaviors against noisy empirical data. MSc Physics of Complex Systems (Universitat de Barcelona), with degrees in Artificial Intelligence, Philosophy of Science, and Sound & Music Computing. Interested in the intersection of Physics and AI.
pc@neocorax.com · github.com/peter-clark · linkedin.com/in/ptrclrk · Barcelona, Spain
Selected Projects
- July 23, 2026
Tapping into Rhythm: How Polyphonic Drum Patterns Map to Monophonic Contours and Support Pattern Retrieval
- Gómez-Marín, D., Clark, P., Jordà, S., Herrera, P. (2026). Music & Science (SAGE). DOI · ORCID 0009-0002-1106-3894
- Studied how listeners compress polyphonic rhythm into a single one-dimensional contour, and which rhythmic features drive it. Continuation of MSc thesis.
- Built the force-sensitive capture system (Python, Pure Data); collected 4,128 taps from 43 participants. Mixed-effects modelling identified frequency-weighted onset density (FWOD) as the dominant predictor.
- Trained neural regressors (PyTorch) mapping contours to coordinates in a 2-D similarity space of 1,513 drum patterns. FWOD best of three representations: R² = 0.79, Spearman ρ = 0.93, 89.6% localised within 10%.
- Ongoing
Interactive complex-systems and physics simulations
,neocorax.com
- Built browser-based statistical-physics simulations, reproducing theoretical thresholds and critical exponents: site & edge percolation recovering the spanning transitions at pc ≈ 0.5927 and pc = 0.5 respectively; random walks showing Brownian motion in multiple dimensions; Bak–Tang–Wiesenfeld sandpile model, self-organizing to ρc = 2.125, avalanches of scale-free nature; forest-fire model implementing the Drossel–Schwabl model showing that fire sizes follow a power-law distribution.
- Engineered the compute path for interactive frame rates on lattices to 256×256 (65,536 sites), ensembles to 20,000 walkers, and a 101-point parameter sweep per realization; union-find with path halving for cluster labelling, multi-source BFS for the shortest crossing, typed arrays throughout, and batched Path2D fills for rendering, resulting in seamless animation on many devices and network speeds.
- Stack — JavaScript (Canvas 2D), Plotly, Jekyll/SCSS, HTML, CSS
- Ongoing
MSc Thesis
,Real World Complex Network Geometry
- Area: hyperbolic geometry of real-world complex networks; embedding and network dynamics.
Technical Skills
- Programming Languages
- Python, C
- Scientific Computing
- NumPy, SciPy, Pandas, numba, NetworkX, SymPy, NetLogo, TensorFlow, PyTorch, JAX
- Skills
- git, Linux, pytest, Docker, CI
Experience
- 02/2021 – Present
Data Analyst / Engineer
,Clark Global Group LLC
- Spain · The Lake Clinic (NGO), Siem Reap, Cambodia
- Produced detailed reports from 1500+ electronic medical records over a 5-year period regarding rural pediatric health spanning 13 villages on a country-wide scale, in alignment with CDC baselines. Presented to local health and NGO professionals highlighting critical public health concerns. Stack SQL, Python, pandas, Matplotlib.
- Developed data processing pipelines reducing time spent on repeat-process tasks for organization of American Physical Therapy Association conference, including delegation of 1000+ paper abstract reviews to relevant reviewers.
- Ran workshop for practical AI-education and integration for company leadership, focused on process standardization, applicability to internal and external organizational gaps.
- 02/2023 – 08/2024
Research Associate
,Multi-Modal Music Interaction Lab
- Universitat Pompeu Fabra, Barcelona, Spain
- Designed and built the data pipeline: direct human–computer interaction capture, processing and analysis with custom-built neural networks, and visual output of experimentally verified music-similarity results to a preexisting navigable interface for music production.
- Extended usability of approach validated in the study above with real-time tap-to-retrieval loop, interpolation within the embedding to generate new patterns, while testing non-discretized real-time input with different machine learning architectures.
- 03/2024 – Ongoing
Scientific Committee Member
,UPF-BMAT Chair on AI and Music
- Universitat Pompeu Fabra, Barcelona, Spain
- Contributed to development and research of policies on algorithmic transparency and responsible evaluation of recommendation systems, aligned with the requirements of the European AI Act.
- 09/2018 – 08/2022
Project Manager
,Rendezvous 050 v.o.f.
- Groningen, Netherlands · Logistics, event organization, executive direction
- Coordinated budget, organization and management of over 15 events with 300+ attendants at multiple venues.
Education
- Current
- MSc Physics of Complex Systems and Biophysics — Universitat de Barcelona
- Relevant keywords: stochastic and noisy processes, complex networks, agent-based modelling, statistical physics, nonlinear dynamics.
- 2023
- MSc Sound and Music Computing — Universitat Pompeu Fabra · 8.8/10
- 2021
- BSc Artificial Intelligence · BA Philosophy of Cognitive Science — University of Groningen
Additional
- Languages
- English (native), Spanish (C1), German (B1), Catalan (A1)
- Work Authorization
- Spanish resident with permission to work; no sponsorship required for EU-based roles.
