sandguine

concordia

A library for generative social simulation

Melting-Pot-Contest-2023

meltingpot

A suite of test scenarios for multi-agent reinforcement learning.

alpaca-lora

Instruct-tune LLaMA on consumer hardware

awesome-model-based-RL

A curated list of awesome model based RL resources (continually updated)

contrastive_metrics

Code for the paper "Learning Temporal Distances: Contrastive Successor Features Can Provide a Metric Structure for Decision-Making"

distributional-sr

Official implementation of the δ-model presented in the paper "A Distributional Analogue to the Successor Representation".

effective-horizon

Code and data for the paper "Bridging RL Theory and Practice with the Effective Horizon"

hanabi.github.io

A list of Hanabi strategies

hidden-context

Code and data for the paper "Understanding Hidden Context in Preference Learning: Consequences for RLHF"

icvf_release

Public code for "Reinforcement Learning from Passive Data via Latent Intentions"

JaxMARL-minimal-information

Multi-Agent Reinforcement Learning with JAX

lab2d

A customisable 2D platform for agent-based AI research

maddpg

Code for the MADDPG algorithm from the paper "Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments"

Mava

🦁 A research-friendly codebase for fast experimentation of multi-agent reinforcement learning in JAX

maxtext

A simple, performant and scalable Jax LLM!

Multi-Agent-Transformer

nanoGPT

The simplest, fastest repository for training/finetuning medium-sized GPTs.

Neural-Network-Zero-to-Hero

Writing keys libraries and core architectures from scratch. Following the tutorials of Neural Network Zero to Hero class from Andrej Karphathy.

overcooked_ai

A benchmark environment for fully cooperative human-AI performance.

paper-reviewer-matcher

Linear programming solver for paper-reviewer matching and mind-matching

pax

Scalable Opponent Shaping Experiments in JAX

purejaxrl

Really Fast End-to-End Jax RL Implementations

pycid

Library for graphical models of decision making, based on pgmpy and networkx

ray

Ray is a unified framework for scaling AI and Python applications. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.

redpoint_hacks

rliable

[NeurIPS'21 Outstanding Paper] Library for reliable evaluation on RL and ML benchmarks, even with only a handful of seeds.

SAELens

Training Sparse Autoencoders on Language Models

value-of-intent

Voyager-Contracts

CAIF