By ML method

These sections are organized by methodology. If you have a physics problem in mind rather than a method, browse by HEP application instead.

Sections

• Equivariant & geometric architectures — symmetry-respecting networks for jets, particles, and detector geometry
• Generative models — normalizing flows, diffusion models, VAEs, GANs in HEP
• Density estimation & likelihood ratios — classifiers as likelihood ratios, reweighting, normalizing flows and diffusion as density estimators
• Uncertainty quantification & calibration — calibrated predictions, conformal methods, Bayesian deep learning
• Explainable AI & interpretability — Shapley values and attribution, symbolic regression, mapping networks onto known observables
• Differentiable programming — end-to-end optimization through simulators and analysis pipelines
• Foundation models — large pre-trained models for collider and detector data
• Agentic workflows — autonomous agents performing tasks

Some sections may be empty for now

The Guide is community-written. A section appears here once a contributor has written it. If a section you care about is missing or only a stub, please consider contributing.