Stars
Official code for RN-D: Discretized Categorical Actors with Regularized Networks for On-Policy Reinforcement Learning (ICML 2026). A drop-in replacement for the Gaussian actor that improves PPO, TR…
LaTeX class file for writing dissertations at UC San Diego
EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL
SPAgent, a foundation agent for understanding, reasoning over, and operating within the physical and spatial world.
Reinforcement Learning of Optimization-Based Control Policies via Implicit Policy Gradients
The code integrates CFD-based airflow simulation, operator learning (neural operator transformer models), and optimization-based control with neural operators to enable energy-efficient and ventila…
A repo for open research on building large reasoning models
[KDD 2025] Awesome Multi-modal Time Series Analysis
This gym provides implementations of various PDEs for easy testing and comparison of data-driven and classical PDE control algorithms.
Python: using OpenStudio Bindings, EnergyPlus API, and matplotlib
A differentiable PDE solving framework for machine learning
The Simulations of Navier-Stokes Equation in 2D and 3D.
My attempt at fluid simulation with the Navier Stokes equations
(EvoApps2022) "Towards a Principled Learning Rate Adaptation for Natural Evolution Strategies"
PythonLinearNonLinearControl is a library implementing the linear and nonlinear control theories in python.
Learning in infinite dimension with neural operators.
Physics Informed Machine Learning Tutorials (Pytorch and Jax)
Pytorch-based framework for solving parametric constrained optimization problems, physics-informed system identification, and parametric model predictive control.
How to train a neural ODE for time series/weather forecasting
Code for our paper Demand Response Model Identification and Behavior Forecast with OptNet: a Gradient-based Approach.
Awesome machine learning for combinatorial optimization papers.
Source code for the dissertation: "Multi-Pass Deep Q-Networks for Reinforcement Learning with Parameterised Action Spaces"