PaperBanana: Automating Academic Illustration For AI Scientists

AirLLM 70B inference with single 4GB GPU

[ICLR 2025] Robust Gymnasium: A Unified Modular Benchmark for Robust Reinforcement Learning.

JMP is a Mixed Precision library for JAX.

Develop your agent for generals.io!

The interface between probabilistic model checking and data-driven policy learning.

A Python implementation of ProVe. ProVe is a formal verifier for safety property for Artificial Neural Networks, presented at the 37th conference on Uncertainty in Artificial Intelligence.

RL environments and tools for spacecraft autonomy research, built on Basilisk. Developed by the AVS Lab.

Github repo for NeurIPS 2024 paper "Safe LoRA: the Silver Lining of Reducing Safety Risks when Fine-tuning Large Language Models"

Differentiable convex optimization layers

LLM inference in C/C++

Real-time webcam demo with SmolVLM and llama.cpp server

Scaling safe exploration to vision control

Code for AAMAS 2024 "Cost-aware Offline Safe Meta Reinforcement Learning with Robust In-Distribution Online Task Adaptation"

(T-IV) Dream to Drive with Predictive Individual World Model

Raster to Vector Graphics Converter

Code for Data Collection & Training in Sim+Real Envs: [RSS 2024] Natural Language Can Help Bridge the Sim2Real Gap

An open-source, code-first Python toolkit for building, evaluating, and deploying sophisticated AI agents with flexibility and control.

An open-source AI agent that brings the power of Grok directly into your terminal.

Minimal JAX implementation unifying Diffusion and Flow Matching algorithms as alternative strategies for transporting data distributions.

Code for our AAAI 26 paper: "Expressive Temporal Specifications for Reward Monitoring"

LOGics formalisms to AUTomata

ICLR 2024: SafeDreamer: Safe Reinforcement Learning with World Models

Official PyTorch implementation of 'Reductive Lie Neurons'

LLM Council works together to answer your hardest questions

dLLM: Simple Diffusion Language Modeling

Learning Safety Constraints for Large Language Models (ICML2025)