Peter Henderson

experiment-impact-tracker. Python

gym-extensions. This repo is intended as an extension for OpenAI Gym for auxiliary tasks (multitask learning, transfer learning, inverse reinforcement learning, etc.)

DeepReinforcementLearningThatMatters. Accompanying code for "Deep Reinforcement Learning that Matters"

PileOfLaw. A dataset for pretraining language models targeted for legal tasks.

DialogDatasets. A repository linking to publicly available dialog datasets. Feel free to send pull requests.

MotionDetection. A project on motion detection in a noisy environment (shaky or moving camera), through background subtraction with single Gaussian models.

OptionGAN. Code accompanying the OptionGAN paper.

echo. Android Mesh Networking Chat with WiFI-Direct

RLSSContinuousControlTutorial. Tutorial on continuous control at Reinforcement Learning Summer School 2017.

EthicsInDialogue. OpenEdge ABL

ReproducibilityInContinuousPolicyGradientMethods. These are experiments for examining reproducibility in Policy Gradient RL algorithms in Continuous domains. Mainly using the Rllab implementation.

MultiStepBootstrappingInRL. Here, we compare Q(\sigma) learning presented by Sutton and Barto in [1] to Tree-Backup, n-step Expected Sarsa, and n-step Sarsa.

SelfDestructingModels. Python

SocraticSwarm. A simulator and algorithms using deccentralized receding horizon control for coordinating autonomous UAV systems in completing a search task.

SarsaVsExpectedSarsa. An a bias-variance tradeoff of Sarsa vs. Expected Sarsa with experiments.

CMACvTileCode. Python

BayesianPolicyGradients. Python

ValuePolicyIterationVariations. Experiments testing variants of Value and Policy iterations.

ExperimentsInIRL. Python

TemporalYolo. Experiments on temporal YOLO

WhatShouldICite. This is an informal record of original citations that I'm aware of for key terms in scientific literature. It started because I didn't know what's the original work to cite for eligibility traces and it seems important to do proper credit assignment.

ClimateChangeFromMachineLearningResearch. Python

orion-pytorch-ppo-acktr-a2c. An adapted version of the ikostrikov RL algorithm implementation for use with the Oríon hyperparameter optimization framework.

AdversarialGain. Python

DeepMultiObjectTracking. Python

drqawrapper. Python

AILawSlides. HTML

TARProtocols. Dataset of Discovery Validation Protocols

NeurIPS. A mirror for some of the NeurIPS website content with a new acronym.

LLM-Tuning-Safety.github.io. CSS

Vulnerabilities-In-Discovery-Tech-Experiment-1. Python

RL-Energy-Leaderboard. Python

echo-laptop. This is the laptop client to to connect to echo nodes

Option-Critic-Turing-Machines. A development toybox and pitch for integrating the option-critic architecture with neural turing machines.