The QLattice is a powerful explainable machine learning technology

Born from the desire to challenge black-box AI, the QLattice is a symbolic AI system that introduces a new standard of explainable AI.

Screenshots of using the QLattice from Python
Screenshots of using the QLattice from Python

What is a QLattice?

The QLattice is a supervised machine learning method for symbolic regression. It is a technology invented by Abzu and inspired by Richard Feynman’s path integral formulation. That’s why we’ve named our Python library Feyn, and the Q in QLattice is for Quantum.

The QLattice is similar to machine learning methods such as Random Forest, Gradient Boosting, and Deep Learning in that it creates predictive models from data.

The QLattice finds explanation graphs using a principle from quantum physics

Explainable and Interpretable

The main advantage of the QLattice is that the models are readily interpretable and explainable. The user is not left with a black-box prediction but also with an explanation for the prediction.

The QLattice won 1st and 2nd place in the recent SRBench Competition 2022: Interpretable Symbolic Regression for Data Science

Less data needed

The QLattice delivers good models and reasonable explanations, even with very little data. This is particularly important in pharma and life sciences, where data is not usually “big”.

Scientific Discovery

The QLattice enables AI-powered scientific discovery. Since the models are simple explanations, they can serve as a scientific hypothesis to be examined further.

Use the QLattice

QLattice – Free For Science

The QLattice is free for non-commercial use. Just “pip install”, and get started with explainable AI through  one of our tutorials


QLattice – Commercial

The QLattice is also available for commercial use in more powerful configurations. Contact us for licensing information


Scientific Service

Our in-house scientists and bioinformaticians can help you get more our of your data, and give you the explainable edge


QLattice Publications

An Approach to Symbolic Regression Using Feyn

Kevin René Broløs, Meera Vieira Machado, Chris Cave, Jaan Kasak, Valdemar Stentoft-Hansen, Victor Galindo Batanero, Tom Jelen, Casper Wilstrup

Symbolic regression outperforms other models for small data sets

Casper Wilstrup, Jaan Kasak