Implementation of deep implicit attention in PyTorch

Biologically-inspired persistent memory engine for Claude Code. 26 cognitive subsystems, Hopfield networks, predictive coding, causal discovery, successor representations, all running locally over SQLite.

Physics-inspired transformer modules based on mean-field dynamics of vector-spin models in JAX

Implementation of approximate free-energy minimization in PyTorch

Create a Hopfield Network for Image Reconstruction

This repository contains the code to reproduce the experiments performed in the Dynamical Mean-Field Theory of Self-Attention Neural Networks article.

The optimisation of the Ising model on various coupling matrices with various methods

Minimum Description Length Hopfield Networks

A Hopfield network to reconstruct patterns (numerical digits) and cope with noise.

Code for Computational Neuroscience course 2020/2021 @ UniPi

Tracing the links between Statistical Mechanics and AI. Phase 1 features a vectorized 2D Ising Model simulation. Phase 2 maps these dynamics to Hopfield Networks to show how physical energy minimization drives memory recall.

The Aetherium Unified Framework: A Topological Gross-Pitaevskii-Poisson Solution for the Dark Fluid Conjecture and Macroscopic Quantum Coherence

A practical comparison between Hopfield Networks and Restricted Boltzmann Machines as content-addressable autoassociative memories.

Learning Similarity Functions in Universal Hopfield Networks

Hopfield networks for pattern recognition

GPU-native equilibrium associative memory: a DEQ-transformer with modern-Hopfield attention, HRR positional binding, and non-autoregressive MaskGIT decoding. Dense matmul + FFT + softmax throughout.

BEAM-Net, a principled neural architecture that unifies three foundational paradigms — modern Hopfield associative memory, Bayesian causal inference, and eventdriven spiking computation — into a single coherent framework with formal convergence guarantees