docs.md

PaperZero Documentation

Reference documentation written by Codex.

Contact me by email or on Slack if you think there are mistakes. You are probably correct.

At A Glance

Topic	Where to Start
API reference by module	Module References
Installation	Setup
Branch responsibilities	Branch Structure
Running Python files correctly	Execution Patterns

Module References

• paper
• framework
• data.origami_sampler
• envs.origami_env

Setup

Use any Python environment manager (venv, Conda, Poetry, etc.).
Run all commands from the repository root.

Install dependencies

python -m pip install -r requirements.txt

Optional: verify imports

python -c "import paper; print('paper ok')"
python -c "from data.origami_sampler import OrigamiSampler; print('sampler ok')"
python -c "from envs.origami_env import OrigamiEnv; print('env ok')"

Expected output:

paper ok
sampler ok
env ok

Branch Structure

Current branch model:

Branch	Purpose
main	Shared integration/docs surface and common infrastructure.
mcst	MCTS-focused development branch.
diffusion	Diffusion-focused development branch.

Use this to choose where feature-specific work should happen.

Execution Patterns

Preferred way to run package files

When running a Python file that imports sibling packages, use module execution:

python -m package.module

Examples:

python -m baselines.train_sl
python -m data.origami_sampler

Why this helps:

• avoids relative import errors from direct path execution
• ensures package resolution is consistent across machines

Direct file execution

Direct execution (python path/to/file.py) might work for standalone scripts, but is less reliable for package-based modules.

Scope

This guide documents the Python import surface and usage patterns. It does not document internal C++ implementation details.