Skip to content

Repository files navigation

world-model

Experiments with world models: agents that learn to predict their environment in latent space — not tokens, not pixels — and use that prediction to act. The working goal is autonomous game-playing agents that learn from their own play, starting with tiny gridworlds and climbing toward real-world transfer (ROADMAP.md).

Quick start

uv sync            # install (Python 3.12, PyTorch, MiniGrid; CUDA used if present, CPU works)
uv run pytest -q   # fast CPU smoke tests

1. See the core problem — and its fix. Naive latent prediction collapses (the encoder cheats by mapping everything to one point); LeJEPA's SIGReg term stops it:

uv run python -m world_model.collect                       # watch the latents collapse
uv run python -m world_model.collect --sigreg-weight 0.05  # SIGReg on — latents survive

2. Train an agent, then watch it play. Train a recurrent world model on DoorKey (partial observability + sparse reward), then let it plan inside the learned latent model — solving the door-then-key task it never saw a reward shaped for:

# train (uses a GPU if available; ~minutes). Produces a checkpoint.
uv run python -m world_model.train_recurrent \
    --env-id MiniGrid-DoorKey-5x5-v0 --rounds 2 --save runs/doorkey.pt

# watch it play in a live window (--record out.gif to save a GIF instead)
uv run python -m world_model.play \
    --checkpoint runs/doorkey.pt --env-id MiniGrid-DoorKey-5x5-v0

Run python -m world_model.play with no --checkpoint for a quick random-agent render on Empty-8x8 (handy to confirm the display works). The experiment behind this path — and how it beats a model-free PPO baseline — is written up in knowledge/experiments/.

What's here

Path What
src/world_model/ envs (Gymnasium/MiniGrid), models (encoder, latent dynamics), agents, replay
tests/ fast CPU smoke tests
knowledge/ curated, agent-maintained research wiki — papers, concepts, ADRs, experiment records
ROADMAP.md the environment ladder
CLAUDE.md / AGENTS.md rules for AI agents working in this repo

The knowledge base

knowledge/ is a three-layer wiki (immutable sources → maintained pages → schema) adapted from Karpathy's LLM Wiki and operated under the curation model of the Context Architecture whitepaper (rendered PDF under the repo's Releases): every page carries status/owner/verified metadata, unread material stays in an intake queue, and learnings from experiments are routed into durable pages instead of dying in chat history. Start at knowledge/INDEX.md.

The whole setup is packaged as a reusable skill in .claude/skills/context-architecture/ — drop it into any project to get the same structure.

License

MIT — see LICENSE.

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages