Ziming Liu

pykan. Kolmogorov Arnold Networks

grow-crystals. Getting crystal-like representations with harmonic loss

BIMT. Brain-Inspired Modular Training (BIMT), a method for making neural networks more modular and interpretable.

Omnigrok. Omnigrok: Grokking Beyond Algorithmic Data

newton-kepler. Understand what physics/algorithms do transformers learn internally when trained on planetary motion

physics_of_skill_learning. We study toy models of skill learning.

schrodinger-pca. Schrodinger Principal Component Analysis

aipoincare. Counting the number of conservation laws from trajectory data

NNPhD-github. Code for "Machine-Learning Non-Conservative Dynamics for New-Physics Detection" (arXiv: 2106.00026)

KindXiaoming.github.io. website

QHMC. Quantum-inspired Hamiltonian Monte Carlo code

sid. discovering interpretable conservation laws from differential equations

aipoincare_2.0. AI Poincare 2.0

entropy-estimator. Jupyter Notebook

ml_hiddensym. Machine Learning hidden symmetries

modded-nanogpt. NanoGPT (124M) in 3 minutes

adam. analyze optimizer properties

stanford-tensorflow-tutorials. This repository contains code examples for the Stanford's course: TensorFlow for Deep Learning Research.

scaling-law-general. Jupyter Notebook

models. Models and examples built with TensorFlow

spatial_transformer_networks. Implementation of spatial transformer networks in keras 2.0 using tensorflow 1.0 as backend.

Expectation_Maximization. expectation maximization

Scaling-law-kerr. Jupyter Notebook

relu_network. Jupyter Notebook