Center for Human-Compatible AI

adversarial-policies

Find best-response to a fixed policy in multi-agent RL

eirli

An Empirical Investigation of Representation Learning for Imitation (EIRLI), NeurIPS'21

evaluating-rewards

Library to compare and evaluate reward functions

human_aware_rl

Code for "On the Utility of Learning about Humans for Human-AI Coordination"

imitation

Clean PyTorch implementations of imitation and reward learning algorithms

overcooked-demo

Web application where humans can play Overcooked with AI agents.

overcooked_ai

A benchmark environment for fully cooperative human-AI performance.

rlsp

Reward Learning by Simulating the Past

seals

Benchmark environments for reward modelling and imitation learning algorithms.

tensor-trust

A prompt injection game to collect data for robust ML research

HumanCompatibleAI/imitation

Clean PyTorch implementations of imitation and reward learning algorithms

HumanCompatibleAI/overcooked_ai

A benchmark environment for fully cooperative human-AI performance.

HumanCompatibleAI/adversarial-policies

Find best-response to a fixed policy in multi-agent RL

HumanCompatibleAI/human_aware_rl

Code for "On the Utility of Learning about Humans for Human-AI Coordination"

HumanCompatibleAI/evaluating-rewards

Library to compare and evaluate reward functions

HumanCompatibleAI/overcooked-demo

Web application where humans can play Overcooked with AI agents.

HumanCompatibleAI/tensor-trust

A prompt injection game to collect data for robust ML research

HumanCompatibleAI/seals

Benchmark environments for reward modelling and imitation learning algorithms.

HumanCompatibleAI/eirli

An Empirical Investigation of Representation Learning for Imitation (EIRLI), NeurIPS'21

HumanCompatibleAI/tensor-trust-data

Dataset for the Tensor Trust project

HumanCompatibleAI/ranking-challenge

Testing ranking algorithms to improve social cohesion

HumanCompatibleAI/leela-interp

Code for "Evidence of Learned Look-Ahead in a Chess-Playing Neural Network"

HumanCompatibleAI/nn-clustering-pytorch

Checking the divisibility of neural networks, and investigating the nature of the pieces networks can be divided into.

HumanCompatibleAI/recon-email

Script for automatically creating the reconnaissance email.

HumanCompatibleAI/reward-preprocessing

Preprocessing reward functions to make them more interpretable

HumanCompatibleAI/multiagent-competition

Code for the paper "Emergent Complexity via Multi-agent Competition"

HumanCompatibleAI/assistance-games

Supporting code for Assistance Games as a Framework paper

HumanCompatibleAI/reducing-exploitability

HumanCompatibleAI/stable-baselines3

PyTorch version of Stable Baselines, improved implementations of reinforcement learning algorithms.

HumanCompatibleAI/dmc2gym

OpenAI Gym wrapper for the DeepMind Control Suite

HumanCompatibleAI/reward-function-interpretability

HumanCompatibleAI/rl-baselines3-zoo

A training framework for Stable Baselines3 reinforcement learning agents, with hyperparameter optimization and pre-trained agents included.

HumanCompatibleAI/sacred

Sacred is a tool to help you configure, organize, log and reproduce experiments developed at IDSIA.

HumanCompatibleAI/katago-driver-bug-repro

Docker files to help reproduce bug described in https://forums.developer.nvidia.com/t/kernel-oops-null-pointer-dereference-when-closing-cuda-application-katago/211270/3

HumanCompatibleAI/pytorch-summary

Model summary in PyTorch similar to `model.summary()` in Keras

HumanCompatibleAI/ranking-challenge-perspective

Prosocial Ranking Challenge Perspective Ranker

HumanCompatibleAI/ray

A fast and simple framework for building and running distributed applications. Ray is packaged with RLlib, a scalable reinforcement learning library, and Tune, a scalable hyperparameter tuning library.

HumanCompatibleAI/rc-submission-civirank

PRC: Civirank submission

HumanCompatibleAI/rc-submission-dante

PRC: Testing ranking algorithms to improve social cohesion

HumanCompatibleAI/sgf-viewer

A simple webpage that can visualize a sgf string encoded as a url fragment.