tedmoskovitz Goto Github PK

bayesian_modeling

A collection of simple Bayesian machine learning methods implemented on toy data.

computational_decipherment

Applying deep learning and other machine learning methods to the decipherment of ancient writing systems.

constrainedrl4lms

A library for constrained RLHF.

convrnn_analysis

Analyze Biologically-Realistic Convolutional Recurrent Networks

deepencoding

deeplearning_thesis

A sample of code from my thesis at Princeton applying deep learning models to neural spike data.

directorv3

Mastering Diverse Domains through World Models

feedback_alignment

Investigating biologically-plausible implementations of the backpropagation algorithm.

first_occupancy

A First Occupancy Representation for Reinforcement Learning

ga_tsp

A simple genetic algorithm (GA) for solving the travelling salesman problem.

lambdarepresentation

Lambda Representation for Diminishing Marginal Utility

minrlhf

A (somewhat) minimal library for finetuning language models with PPO on human feedback.

nanogptconstraints

pmpp

notes and code for Programming Massively Parallel Processors, 4th Edition

practicecpp

Simple C++ Programs

ram_cifar10

reinforcement_learning

My solutions to Denny Britz's short course on RL.

simple_database

A simple relational database.

simplecuda

Simple Neural Network in CUDA

simpleppo

A Simple, Easily-Customizable, Fully Jitted PPO Implementation in Jax

svgd

Stein Variational Gradient Descent

tdsr_python

successor representation for RL

top

Implementation of Tactical Optimistic and Pessimistic value estimation

tvpo

An implementation of Total Variation Policy Optimization (TVPO)

utils

wnpg

implementation of Wasserstein Natural Policy Gradients and Wasserstein Natural Evolution Strategies