A self-improving and iterating agent harness. Tell the agent what your goal is: it does deep-research, writes a spec, iterates, self-refines, and works toward a goal.

GTNH Factory Flow planner and dataset automation

A library for mechanistic interpretability of GPT-style language models

An open infrastructure to democratize and decentralize the development of superintelligence for humanity.

Train the smallest LM you can that fits in 16MB. Best model wins!

code for the paper "Stein Variational Gradient Descent (SVGD): A General Purpose Bayesian Inference Algorithm"

AI agents running research on single-GPU nanochat training automatically

The missing DevTools for Claude Code — inspect session logs, tool calls, token usage, subagents, and context window in a visual UI. Free, open source.

An open-source attempt at training a variant of LeCun's energy-based models (EBM) to reason in latent space and solve Sudoku.

a glimpse into the future of eye tracking in terminals

🌳 MCTS-inspired parallel beam search for conversation optimization. Explore multiple dialogue strategies simultaneously, stress-test against diverse user personas, score with multi-judge consensus…

MCP-enabled AI conversation engine with MCTS analysis, FastAPI backend, and async operations for building advanced LLM applications