DSPy: The framework for programming—not prompting—foundation models

Large Language Model based Multi-Agents: A Survey of Progress and Challenges

Training Sparse Autoencoders on Language Models

Python/PyMC3 port of the examples in " Statistical Rethinking A Bayesian Course with Examples in R and Stan" by Richard McElreath

A library for generative social simulation

Official repository for my MSc thesis: "Addressing Goal Misgeneralization with Natural Language Interfaces."

Inspect: A framework for large language model evaluations

The multiagent extension for the PDDL parser

METR Task Standard

This repository contains the jailbreaking process for GPT-3, GPT-4, GPT-3.5, ChatGPT, and ChatGPT Plus. By following the instructions in this repository, you will be able to gain access to the inne…

ChatGPT DAN, Jailbreaks prompt

A library for generative social simulation

Unofficial Notion.so app for Linux

Scalable Opponent Shaping Experiments in JAX

The Spotify Data Analysis Project showcases data's role in diverse fields, using Python and libraries like Pandas,Numpy,Seaborn and Matplotlib, within the Jupyter Notebook environment. It explores …

Minor mode for Emacs to improve English writing

Simplified model for reasoning about agents maximizing their expected utilities

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

A tool for aggregating and plotting MARL experiment data.

🕹️ A diverse suite of scalable reinforcement learning environments in JAX

⚡ Flashbax: Accelerated Replay Buffers in JAX

Datasets with baselines for offline multi-agent reinforcement learning.

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

Code for Model-Free Opponent Shaping (ICML 2022)

Minimal implementation of clipped objective Proximal Policy Optimization (PPO) in PyTorch

Accompanying code for the paper: On the Impossibility of Global Convergence in Multi-Loss Optimization.

Notes for Judea Pearl et al., *Causal Inference in Statistics, a Primer*

Latex package for drawing (causal) influence diagrams (CIDs) and labeling incentives, maintained by the Causal Incentives Working Group