Scientists use generative AI to answer complex physics questions

This article was reviewed according to Science

fact checked

peer-reviewed publication

trusted source


Credit: Pixabay/CC0 public domain

× close to

Credit: Pixabay/CC0 public domain

When water freezes, it changes from a liquid phase to a solid phase, resulting in a drastic change in properties such as density and volume. Phase transitions in water are so common that most of us probably don’t even think about them, but phase transitions in new materials or complex physical systems are an important area of ​​research.

To fully understand these systems, scientists must be able to recognize phases and detect the transitions between them. But quantifying phase changes in an unknown system is often unclear, especially when data is sparse.

Researchers from MIT and the University of Basel in Switzerland applied generative artificial intelligence models to this problem and developed a new machine-learning framework that can automatically map phase diagrams for new physical systems.

Their physics-based approach to machine learning is more efficient than laborious, manual techniques that rely on theoretical expertise. Importantly, because their approach uses generative models, it does not require the massive, labeled training datasets used in other machine learning techniques.

Such a framework could help scientists investigate the thermodynamic properties of new materials or, for example, detect entanglement in quantum systems. Ultimately, this technique could allow scientists to autonomously discover unknown phases of matter.

“If you have a new system with completely unknown properties, how would you choose which observable quantity to study? The hope is, at least with data-driven tools, that you can scan large new systems in an automated way, and that will indicate alert you to important changes in the system.

“This could be a tool in the pipeline of automated scientific discovery of new, exotic properties of phases,” says Frank Schäfer, a postdoc in the Julia Lab in the Computer Science and Artificial Intelligence Laboratory (CSAIL) and co-author of a article about this approach.

Schäfer is joined on this paper by first author Julian Arnold, a graduate student at the University of Basel; Alan Edelman, professor of applied mathematics in the Department of Mathematics and leader of the Julia Lab; and senior author Christoph Bruder, professor at the Department of Physics at the University of Basel.

The research was published in Physical Assessment Letters.

Detect phase transitions using AI

While the transition from water to ice is perhaps one of the most obvious examples of a phase change, more exotic phase changes, such as when a material transitions from a normal conductor to a superconductor, are of great interest to scientists.

These transitions can be detected by identifying an ‘order parameter’, a quantity that is important and expected to change. For example, water freezes and turns into a solid phase (ice) when the temperature drops below 0°C. In this case, an appropriate order parameter could be defined in terms of the proportion of water molecules that are part of the crystalline lattice versus the proportions that remain in a disordered state.

In the past, researchers relied on physics expertise to manually build phase diagrams, relying on theoretical insight to know which order parameters are important. Not only is this annoying for complex systems, and perhaps impossible for unknown systems with new behavior, but it also introduces human biases into the solution.

More recently, researchers have begun to use machine learning to build discriminative classifiers that can solve this task by learning to classify a metric as coming from a particular phase of the physical system, in the same way that such models classify an image as a cat or a dog.

The MIT researchers have shown how generative models can be used to solve this classification task much more efficiently and in a physics-informed manner.

The Julia Programming Language, a popular language for scientific computing that is also used in MIT’s introductory linear algebra classes, offers many tools that make it invaluable for constructing such generative models, Schäfer adds.

Generative models, such as the models underlying ChatGPT and Dall-E, typically work by estimating the probability distribution of some data, which they use to generate new data points that fit the distribution (such as new cat images that are similar to existing cat images). .

However, when simulations of a physical system using proven scientific techniques are available, researchers are provided with a model of its probability distribution for free. This distribution describes the metrics of the physical system.

A better informed model

The MIT team’s insight is that this probability distribution also defines a generative model on which a classifier can be constructed. They plug the generative model into standard statistical formulas to directly construct a classifier instead of learning it from samples as was done in discriminative approaches.

“This is a really neat way to incorporate something you know about your physical system deep into your machine learning scheme. It goes far beyond just doing feature engineering on your data samples or simple inductive biases,” says Schäfer.

This generative classifier can determine which phase the system is in, given a certain parameter, such as temperature or pressure. And because researchers approach the probability distributions underlying measurements directly from the physical system, the classifier has system knowledge.

This makes their method outperform other machine learning techniques. And because it can operate automatically without the need for extensive training, their approach significantly improves computational efficiency in identifying phase transitions.

At the end of the day, similar to how you would ask ChatGPT to solve a math problem, the researchers can ask the generative classifier questions such as “does this sample belong to Phase I or Phase II?” or “was this sample generated at high or low temperature?”

Scientists could also use this approach to solve various binary classification tasks in physical systems, possibly to detect entanglement in quantum systems (is the state entangled or not?) or to determine whether theory A or B is best suited to solve a problem. to solve a particular problem. They could also use this approach to better understand and improve large language models like ChatGPT by identifying how to tune certain parameters so that the chatbot gives the best results.

In the future, the researchers also want to study theoretical guarantees about how many measurements they would need to effectively detect phase transitions and estimate the amount of calculations required to do so.

More information:
Julian Arnold et al., Mapping phase diagrams with generative classifiers, Physical Assessment Letters (2024). DOI: 10.1103/PhysRevLett.132.207301. On arXiv (2023): DOI: 10.48550/arxiv.2306.14894

Magazine information:
Physical Assessment Letters


Leave a Reply

Your email address will not be published. Required fields are marked *