The agent engineering platform.

Massively Accelerated Learning of Operators for Quantum-transport

Python package to aid materials design and informatics

RetroAgent: Harnessing LLMs to Search Over Structured Memory for Agentic Retrosynthesis Planning

Simple Guidance Mechanisms for Discrete Diffusion Models

Code for the paper https://arxiv.org/abs/2402.04997

Implementation for SE(3) diffusion model with application to protein backbone generation

Functional Attention: From Pairwise Affinities to Functional Correspondences

Torch-native C++/CUDA library to accelerate tensor-product layers in MLIPs

OpenEquivariance: a fast, open-source GPU JIT kernel generator for the Clebsch-Gordon Tensor Product.

cuEquivariance is a math library that is a collective of low-level primitives and tensor ops to accelerate widely-used models, like DiffDock, MACE, Allegro and NEQUIP, based on equivariant neural n…

Claude Code skills, subagents, and slash commands for scientific research (paper writing, figures, document automation, Korean research paperwork)

Slowly annealed Langevin dynamics

code for "Riemannian Flow Matching on General Geometries".

Optimize prompts, code, and more with AI-powered Reflective Optimization

Differentiable generalized eigensolver in JAX. Extracted from PYSCFAD implementation.

Cuda kernels for leveraging LLM sparsity to improve throughput and decrease the memory requirements during inference and training.

Skala exchange-correlation functional

Local-first code intelligence graph for MCP and CLI. Builds a persistent map of your codebase so AI coding tools read only what matters, with benchmarked context reductions on reviews and large-rep…

The official implementation of 'Equivariant Denoisers Cannot Copy Graphs: Aligned your Graph Diffusion Models' (ICLR2025)

MARA is a plug-and-play SE(3)-equivariant angular-radial attention module that improves accuracy and robustness of molecular force fields.

Orchestration layer for coding agents (Claude Code, Codex)

AI agents running research on single-GPU nanochat training automatically

Teams-first Multi-agent orchestration for Claude Code

🤗 ml-intern: an open-source ML engineer that reads papers, trains models, and ships ML models

Pytorch implementation of MeanFlow on ImageNet and CIFAR10