Author & Independent Researcher
Books and research exploring intelligence, mathematics and the structures behind how things work.
01 / The writing
Complex ideas.
A closer reading.
Forthcoming
What really happens when a system learns?
The book offers a structural view of machine learning, moving from data and patterns to transfer, invariants, state spaces, and the testing of stable structure.
Its central question is simple: did the system truly discover structure, or did it merely learn to reproduce familiar conditions well?
Learning begins with experience, but structure begins with what survives change.
Future books and announcements will be collected here.
02 / In progress
Books, essays and ongoing intellectual projects exploring intelligence, adaptation, mathematics and the structural foundations of complex systems.
Independent investigations into mathematical structures, system behaviour, cognitive processes and the conditions that make complex phenomena possible.
Research notes, mathematical models, diagrams, experimental materials and technical appendices supporting published and ongoing work.
03 / Behind the ideas
The books explain the ideas. Research offers a closer look at the models, arguments and experiments behind them.
Explore on GitHub04 / The author
Alex Marin is the pen name of Alexey Gudilin, an independent researcher, author, and founder of Nelvarx Systems AB, based in Sweden.
His research focuses on intelligence, mathematical modelling, complex systems, structural analysis, and adaptation. Particular attention is given to the conditions under which systems maintain stability, undergo transformation, learn, and develop new capabilities.
At the centre of this work is an effort to understand not only how systems function, but why particular processes and transformations become possible in the first place. This approach brings together mathematical models, theoretical investigations, and the development of practical analytical methods.
The books published under the name Alex Marin form a distinct part of this work. Their purpose is to make complex ideas accessible to a wider audience while maintaining a connection to mathematical foundations, research materials, and verifiable reasoning.
Selected research and supporting materials are made available through GitHub. New work is released progressively as it is prepared and reviewed.