A simpler version of my dotfiles that I made for people at work.

MapleStory Client built on Wasm playable on the web

BullshitBench measures whether AI models challenge nonsensical prompts instead of confidently answering them, created by Peter Gostev.

Post-training with Tinker

James' cookbook of evaluations and finetuning experiments

Weird Generalization and Inductive Backdoors Website

An alignment auditing agent capable of quickly exploring alignment hypothesis

Code and materials for "Weird Generalization and Inductive Backdoors"

Atropos is a Language Model Reinforcement Learning Environments framework for collecting and evaluating LLM trajectories through diverse environments

Simple RL training for reasoning

A python sdk for LLM finetuning and inference on runpod infrastructure

AI Agent Framework, the Pydantic way

A coding agent for open models like Kimi K3

Improving transparency of large language models' reasoning

Moonshot - A simple and modular tool to evaluate and red-team any LLM application.

Code for ICLR 2025 Failures to Find Transferable Image Jailbreaks Between Vision-Language Models

datasets from the paper "Towards Understanding Sycophancy in Language Models"

Inspect: A framework for large language model evaluations

Adding guardrails to large language models.

Embedding Markdown in a Blogger HTML page.

Alpaca dataset from Stanford, cleaned and curated

Laaxus Revision of AI, a Victoria 3 mod

Representation Engineering: A Top-Down Approach to AI Transparency

the LLM vulnerability scanner

A reactive async streaming library for asyncio / trio!

GitHub checkout action with LFS files pulled from cache

GitHub Action for pyright