Inverse Design Python Packages

beamz

GPU-accelerated FDTD Python framework for 2D/3D electromagnetic simulation and photonic inverse design.

fabtwin

Learned generative process twins for yield-aware inverse design of multilayer optics: exact TMM with a hand-derived adjoint, pathwise CVaR robustification through twin and physics

splayout

Silicon Photonics Design Tools.

torchoptics

Differentiable wave optics simulation library built on PyTorch

rfx-fdtd

JAX-based 3D FDTD electromagnetic simulator

delfin-complat

Computational Chemistry Meets Agentic Machine Learning

pycharge

Electromagnetics simulation library for moving point charges built on JAX

difftmm

Differentiable Transfer-Matrix Method thin film solver.

torch-molecule

torch-molecule is a deep learning package for molecular discovery, designed with an sklearn-style interface for property prediction, inverse design and representation learning.

canonical-sets

Exposing Algorithmic Bias with Canonical Sets

pen-stack

A verification and grounding layer for AI-driven genome writing. Foundation models and a grounded co-scientist agent propose edits; PEN-STACK checks each against validated tools, never a language model: where a write is safe and durable, which enzyme makes it, and where it goes off-target. Calibrated, provenance-tracked, pre-registered.

mst-topopt

Topology optimization (TopOpt) framework based on the Maxwell stress tensor (MST) formalism to optimize for optical force-dependent figures of merit.

bcp-perception

Nested Resonance Memory (NRM) by Aldrin Payopay — a reality-compiler research archive. Budget-Constrained Perception (V = G − λC), the Information↔Matter isomorphism, inverse cymatics, emergence experiments, and Duality-Zero / Helios engineering.

glas

Genetic Lookup for Apt Substances

katmer

katmer is a powerful library for optimizing the design of optical thin films using automatic differentiation via JAX and Equinox, enabling efficient and accurate inverse design solutions.