bandit

Bandit is a tool designed to find common security issues in Python code.

Performing security tests inside your CI

Automated security testing using bandit and flake8.

Python boilerplate using uv, pre-commit, prettier, pytest, GitHub Actions, mypy, ruff, bandit & docformatter.

Python application to setup and run streaming (contextual) bandit experiments.

Contextual Bandits in R - simulation and evaluation of Multi-Armed Bandit Policies

Thompson Sampling Tutorial

GitHub Action to run Bandit

A pre-commit hook to find common security issues in your Python code

Code for Paper (Policy Optimization in RLHF: The Impact of Out-of-preference Data)

Another A/B test library

Frontend to display data from huskyCI analyses

OverTheWire wargames, structured walkthroughs written to go through, not just solve.

[NeurIPS 2023 Spotlight] In-Context Impersonation Reveals Large Language Models' Strengths and Biases

github action to run the bandit security linter

pytest plugin to execute bandit across a codebase

This repository contains all public UVR models, with possibly a few private ones included.

An official JAX-based code for our NeuraLCB paper, "Offline Neural Contextual Bandits: Pessimism, Optimization and Generalization", ICLR 2022.

Flask Project Made with Poetry

We use policy gradient to help agents learn optimal policies in a competitive multi-agent contextual bandit setting