PIK GaNe's Projects
Modeling agents with probabilistic programs
This is a suite of reinforcement learning environments illustrating various safety properties of intelligent agents.
A validated automatic evaluator for instruction-following language models. High-quality, cheap, and fast.
A simulation framework for RLHF and alternatives. Develop your RLHF method without collecting human data.
Code for reproducing the results from the paper Avoiding Side Effects in Complex Environments
High-quality single file implementation of Deep Reinforcement Learning algorithms with research-friendly features (PPO, DQN, C51, DDPG, TD3, SAC, PPG)
Official codebase for Decision Transformer: Reinforcement Learning via Sequence Modeling.
Simple and easily configurable grid world environments for reinforcement learning
python package for torch-based neural network version of MoTaBaR
Markov Decision Process (MDP) Toolbox for Python
Python framework for optimization of epidemic testing strategies
quantify agents' degrees of moral responsibility in complex multi-agent decision situations
Reinforcement Learning through Active Inference with additional safety measures
A modular RL library to fine-tune language models to human preferences
Satisficing-based Intelligent Agents
A repo to explore multi-agent reinforcement learning in the context of aspiration based, non-maximising agents. This project is part of the Supervised Program for Alignment Research.
Contrib package for Stable-Baselines3 - Experimental reinforcement learning (RL) code
PyTorch version of Stable Baselines, reliable implementations of reinforcement learning algorithms.
Pytorch implementation on OpenAI's Procgen ppo-baseline, built from scratch.
TriCl model in C++
We develop an interactive, consensus-oriented group decision app
Webppl library for generating Gridworld MDPs. JS library for displaying Gridworld. Additional agents that satisfice.