# ALIGNN Projects at AI Tinkerers

> Canonical HTML: https://aitinkerers.org/technologies/alignn
> Markdown URL: https://aitinkerers.org/technologies/alignn.md
> Technology record last updated: 2026-05-26T00:00:41Z
> Generated: 2026-09-20T17:35:53Z

ALIGNN is a deep learning framework that models both bond distances and angles to predict material properties with high precision.

Developed by researchers at NIST, the Atomistic Line Graph Neural Network (ALIGNN) solves a major limitation in materials modeling: capturing the exact geometry of atomic structures (1.2.2). By using a dual-graph architecture, the model alternates message passing between an interatomic bond graph and its corresponding line graph (1.2.3). This allows ALIGNN to explicitly track both two-body (bond length) and three-body (bond angle) interactions (1.1.2). The framework outperforms traditional graph neural networks by up to 85% in accuracy, making it a highly reliable tool for predicting over 100 physical properties (including band gaps and elastic moduli) across 89 elements (1.2.7, 1.2.8).

- Official technology site: https://github.com/usnistgov/alignn
- Public AI Tinkerers demos and talks: 1
- Result page: 1 of 1

## Recent Public Talks and Demos

### [Transformer-Diffusion model for molecular battery material generation](https://dubai.aitinkerers.org/talks/rsvp_SK0KgeEQ2Rc)

I built Simer Energy, an end-to-end generative AI pipeline that uses a hybrid Transformer-Diffusion architecture to autonomously design, physically relax and thermodynamically validate novel battery materials from scratch. For the demo, I will execute a terminal-based run of the by inputting strict material constraints (e.g., elemental bounds for a cobalt-free transition metal oxide, target stoichiometry, and symmetry parameters) into a fine-tuned model. I’ll show how the Transformer maps these constraints into a discretespace groups and passes it as a conditioning vector to a diffusion model. You will see the model denoise the 3D spatial atomic coordinates, followed by the critical "zero-strain" and then passing the raw generated structure directly into a Universal Machine Learning Force Field (CHGNet) to instantly relax the atomic coordinates, bypassing days of expensive DFT calculations and then run the relaxed structure through ALIGNN to validate its Energy Above Hull, resulting in a mathematically viable file generated in under three minutes.

- Event context: AI Tinkerers Dubai - May Demo Day — 2026-05-23 — Dubai
- Public talk page: https://dubai.aitinkerers.org/talks/rsvp_SK0KgeEQ2Rc

## Related Technologies

- [Amazon SageMaker](https://aitinkerers.org/technologies/amazon-sagemaker) ([Markdown](https://aitinkerers.org/technologies/amazon-sagemaker.md)) — 2 public demos
- [AWS SageMaker](https://aitinkerers.org/technologies/aws-sagemaker) ([Markdown](https://aitinkerers.org/technologies/aws-sagemaker.md)) — 2 public demos
- [CDVAE](https://aitinkerers.org/technologies/cdvae) ([Markdown](https://aitinkerers.org/technologies/cdvae.md)) — 1 public demo
- [CHGNet](https://aitinkerers.org/technologies/chgnet) ([Markdown](https://aitinkerers.org/technologies/chgnet.md)) — 1 public demo
- [Hugging Face](https://aitinkerers.org/technologies/hugging-face) ([Markdown](https://aitinkerers.org/technologies/hugging-face.md)) — 41 public demos
- [Llama 3](https://aitinkerers.org/technologies/llama-3) ([Markdown](https://aitinkerers.org/technologies/llama-3.md)) — 38 public demos
- [Meta Llama 3](https://aitinkerers.org/technologies/meta-llama-3) ([Markdown](https://aitinkerers.org/technologies/meta-llama-3.md)) — 1 public demo
- [NVIDIA A10G](https://aitinkerers.org/technologies/nvidia-a10g) ([Markdown](https://aitinkerers.org/technologies/nvidia-a10g.md)) — 1 public demo
- [PyTorch](https://aitinkerers.org/technologies/pytorch) ([Markdown](https://aitinkerers.org/technologies/pytorch.md)) — 273 public demos
- [unsloth](https://aitinkerers.org/technologies/unsloth) ([Markdown](https://aitinkerers.org/technologies/unsloth.md)) — 8 public demos