The goal of the IAIFI is to advance knowledge of fundamental interactions from the smallest to the largest scales using innovative methods in AI built upon ab initio physics principles, while simultaneously advancing the foundations of AI. Specifically, we are targeting opportunities for ab initio AI to improve theory calculations, to improve experiments, and to advance the field of AI. While our work includes a broad spectrum of physics and AI applications, our core focus is developing the nascent field of ab initio AI itself.

Coming soon: A full set of IAIFI project pages, each with more details about our research work, publications, etc.

Physics Theory

For many physics problems, the governing equations that encode the fundamental physical laws are known. Undertaking key calculations within these frameworks, as is essential to test our understanding of the universe and guide physics discovery, however, can be computationally demanding or even intractable. A primary research area for the IAIFI is AI for such ab initio theory studies. AI for first-principles physics naturally requires approaches with specific features, such as guarantees of exactness, reproducibility, and robust handling of uncertainties.

The specific ab initio physics approaches that we are targeting include first-principles calculations within the Standard Model (SM) of nuclear and particle physics, work to understand physics beyond the SM in the framework of string theory, and theory calculations for astroparticle physics. Complementing these efforts is work towards approaches that do not use AI to study a known theory, but instead to determine what previously unknown theory underlies observations. While the target physics spans a wide range of frameworks and scales, success in these areas all rely intimately on our efforts to advance the field of AI itself with work in interpretability, speed, and incorporating physical symmetries into AI frameworks, i.e. to develop the field of ab initio AI.

Research Highlights

IAIFI Physics Theory Projects: Accelerating Lattice Field Theory with AI, Exploring the Multiverse with AI, Classifying Knots with AI, Astrophysical Simulations with AI, Toward an AI Physicist, and String Theory Conjectures via AI.

Physics Experiment

By developing the field of ab initio AI, the IAIFI is also impacting many experimental applications, e.g., where ab initio principles inform the design of AI methods that are more easily verifiable using well-understood calibration data samples, leading to better quantification of uncertainties. We are working to improve the operations of and enhance the physics potential of the Large Hadron Collider (LHC), the Laser Interferometer Gravity Wave Observatory (LIGO), and astrophysical experiments relevant for the nascent field of multi-messenger astronomy. Our efforts in ab initio AI for experiment are symbiotic with our work to advance the field of AI itself, particularly with the focus on interpretability and speed. Finally, interpreting experimental results often requires fast model evaluation or simulation; improving these tasks naturally has much in common with our efforts to accelerate or make tractable ab initio physics theory calculations.

Research Highlights

IAIFI Physics Experiment Projects: Interpretable AI for the LHC, Real-Time Processing with AI, Neutrinos Imaging with AI, LIGO Noise Reduction with AI, Accelerating Gravitational Waveform Computations with AI, and Identifying Electromagnetic Counterparts for Multi-Messenger Astrophysics with AI.

Foundational AI

As discussed above, ab initio AI promises to greatly improve theoretical and experimental physics in ways that we are only beginning to explore. In turn, the unique features of these physics applications offer compelling research opportunities in AI more broadly. For example, physical systems often feature exact or approximate symmetries. Exploiting these known invariances in data structures and algorithms is a key to developing efficient algorithms for physics applications, but is also likely to be transformative in broader contexts. Physics-informed architectures and hardware development promise advances in the speed of AI algorithms, and work in statistical physics is providing a theoretical foundation for understanding AI dynamics. The IAIFI is supporting these efforts that deeply entwine our ab initio AI research with our ab inito physics goals.

Research Highlights

IAIFI AI Projects: Embedding Quantum Field Theory Principles into AI, AI that Respects Algebraic and Geometric Invariance, Fast AI on Specialized Hardware, AI at the Speed of Light, Speeding up Simulations with Curiosity-Driven AI, From Sequential AI to Parallel AI, Statistical Physics for Sparse AI Learning, Combining Symmetries and Flexibility with AI, AI for Data/Simulation Differences, and Statistical Physics for AI Generalization.


IAIFI researchers are incorporating complex ab initio principles into AI research to revolutionize how AI is used in the study of fundamental interactions, while simultaneously advancing the foundations of AI. Join us!