Agentic RL Training at Scale

🟣 Concurrency interview questions and answers to help you prepare for your next software architecture and design patterns interview in 2026.

🔥 LeetCode for PyTorch — practice implementing softmax, attention, GPT-2 and more from scratch with instant auto-grading. Jupyter-based, self-hosted or try online.

Supplementary material for research on investigating model incrimination

Mobile and Web client for Codex and Claude Code, with realtime voice, encryption and fully featured

An alignment auditing agent capable of quickly exploring alignment hypothesis

This was designed for interp researchers who want to do research on or with interp agents to give quality of life improvements and fix some of the annoying things you get from only using Claude cod…

A framework for testing and evaluating AI agents across various task domains, designed for misalignment interpretability research.

Experimental framework for studying AI agent behavior in game environments. Analyzes decision-making patterns, hint usage, and reward hacking through controlled rollouts and causal analysis.

Reward hacking benchmark

aider is AI pair programming in your terminal

Chess agent specification gaming

🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.

Parameter Decomposition

Real-Time Detection of Hallucinated Entities in Long-Form Generation

Tokamax: A GPU and TPU kernel library.

TPU inference for vLLM, with unified JAX and PyTorch support.

The nnsight package enables interpreting and manipulating the internals of deep learned models.

Ollama Python library

Open-source implementation of AlphaEvolve

This project provides resilient training logic to improve the training efficiency of large-scale distributed workloads on Google Cloud.

Recipes for reproducing training and serving benchmarks for large machine learning models using GPUs on Google Cloud.