Structural AI Reasoning
Investigating alternatives to autoregressive generation — models that refine holistic structures rather than sequential tokens. Exploring implications for scientific reasoning and complex system modeling.
View on GitHubI'm an aerospace engineering student at New Uzbekistan University, drawn to the boundary between what we know and what we can build. My interests span quantum mechanics, mathematical physics, artificial intelligence, and computational simulation.
I haven't narrowed down yet — and I see that as a strength. The hardest problems live at the intersections. I learn by building: formulate a question, model it, simulate it, break it, and iterate.
Currently seeking Research Assistant positions where I can contribute to ambitious scientific questions and grow as a researcher.
Investigating alternatives to autoregressive generation — models that refine holistic structures rather than sequential tokens. Exploring implications for scientific reasoning and complex system modeling.
View on GitHubExperiments in generating modular neural expertise attachable to base language models. Focused on efficient specialization without catastrophic forgetting.
View on GitHubComputational experiments modeling physical systems and mathematical frameworks. Validating theoretical predictions through simulation before physical prototyping.
View on GitHubSmall-scale prototypes designed around ordinary problems — with emphasis on making something dramatically simpler, smarter, or more useful through first-principles thinking.
View on GitHubFundamental limits of measurement, superposition, and information.
Novel architectures, reasoning beyond autoregression, emergent capability.
Geometry of spacetime, symmetry groups, and physical law.
Structures, propulsion, and the discipline of flight.
Computational experiments as a bridge between theory and prototype.
Strange questions that deserve a test, not just a discussion.