About
This blog is where I share my research and thoughts on artificial intelligence. The goal isn’t to track the latest tools or hype, but to dig deeper into theory, methods, and sometimes into entirely new ways of thinking about intelligent systems.
Each post is an attempt to make sense of complex ideas without stripping away their depth. I’ll borrow from academic research, industry practice, and my own experiments, but always with the same purpose: to build and test ideas that might expand how we understand intelligence.
This space is meant for researchers, professionals, and curious readers who care about clarity and precision. It’s not just commentary, it’s a place to propose, challenge, and refine concepts that could push the field forward.
In short, this is a working log of exploration. A record of trying to move past what exists today, toward what might come next.
Blog
Gradient flow in Laplace Neural Network
This post breaks down how gradients propagate through layers in Laplace neural network and how to avoid vanishing and exploding gradients. With a focus on the equations behind backpropagation in Laplace domain, it offers a clear, research-driven explanation of how learning is sustained inside Laplace neural networks.
How to model the world? Introduction to Laplace Neuron
What does it really mean to “model a world”? In artificial intelligence, this question goes beyond data and algorithms, it touches on how machines can represent, reason, and adapt to dynamic environments. This post explores fresh perspectives on world modeling, offering insights into why it matters for the future of intelligent systems and where current approaches fall short.