Verified Deep Learning with Lean 4
Formally verifies ResNet and ViT architectures using Lean 4 proofs.

Educational content in a space where Nathan Lambert's RLHF book already exists.
ML researchers and engineers working on RLHF and agent training
Nathan Lambert's RLHF Book · Sutton & Barto
Formally verifies ResNet and ViT architectures using Lean 4 proofs.
Research scripts testing reward governance thesis with no product surface.
Teaches LLM RL training with working Tic Tac Toe demo that beats gpt-5-mini.
Revives deprecated OpenAI gym-http-api with Docker images and built-in browser monitoring views.
Ed25519 signature verification in browser solves agent accountability for disputes.
Deterministic offline tamper detection—pinned at capture, replayed without side effects.