Skip to content

AGENTS.md

Latest commit

 

History

History
57 lines (46 loc) · 2.29 KB

AGENTS.md

File metadata and controls

57 lines (46 loc) · 2.29 KB
Raw

Methods AI Assistant — Statistical & Computational Methods Wiki

This sub-wiki gives an AI assistant the scientific context of the methodological tooling that underpins PyAutoLens, PyAutoFit, and related codes: Bayesian inference, nested and Hamiltonian sampling, deep learning for astrophysics, probabilistic programming languages, linear-algebra primitives (NUFFT, message passing), and the wider scientific-software ecosystem. Sibling of ../lensing/ under PyAutoMemory/, same Karpathy "LLM Wiki" pattern.

Layout

methods/
├── AGENTS.md           # this file — schema + scope
├── index.md            # top-level navigation
├── log.md              # append-only compilation log
├── concepts/           # Bayes basics, samplers, deep learning, etc.
├── entities/           # PyAutoFit, PolyChord, BlackJAX, jax-finufft, …
├── sources/            # per-topic claim support
└── seed/               # unverified 2026-05 import — never cite from here

Schema

Inherits verbatim from ../AGENTS.md.

Scope

In-scope folders:

  • Stats/ — Bayesian methods, likelihood-free inference, MCMC, sampler comparisons.
  • GaussianLinearModels/ — Gaussian linear models, conjugate priors, closed-form posteriors used in linear inversion.
  • PPLs/ — probabilistic programming languages (Stan, PyMC, NumPyro, Pyro, etc.).
  • Deep Learning/ — CNNs, BNNs, simulation-based inference applied to astrophysics (cross-references ../lensing/sources-deep-learning-lensing.md).
  • Software/ — general-purpose scientific software references (samplers, optimisers, JAX ecosystem).
  • Simulation/ — Monte Carlo simulation methodology, hierarchical forward modelling.
  • Root-level PolyChord.pdf, linearalgebra.pdf, NUFFTPAper.pdf, PYNUFFT.pdf, MessagePassingIsAWayOfLife.pdf, BigData.pdf, PyLops (folder), imfit.pdf, Profit2019.pdf.

Adjacent topics that link out:

  • Lens-modelling-specific Bayesian / regularisation methods — ../lensing/concepts/bayesian-inference-lensing.md and ../lensing/concepts/regularization.md.
  • Mass-model / source-reconstruction software entities — ../lensing/entities/{pyautolens,pyautofit,lenstronomy}.md.

How the assistant should use this wiki

Same protocol as ../AGENTS.md.