BIDMaP is happy to be jointly hosting this year's seminar series with colleagues in computational physical sciences, including collaborators from the departments of Physics and Astronomy, and the Lawrence Berkeley National Laboratory.
Understanding how inorganic materials and minerals behave, evolve, and change in various environmental conditions is a core challenge in geophysical sciences, and an area where ab-initio simulation methods have long been helpful. The same processes can also be used to find and discover materials for...
Deep learning techniques are increasingly applied to scientific problems where the precision of networks is crucial. Despite being deemed as universal function approximators, neural networks, in practice, struggle to reduce prediction errors below a certain threshold even with large network sizes and extended training iterations...
We consider the following pervasive scenario in contemporary scientific discovery: we are faced with an enormous space of candidate designs for some task we must explore. However, our ability to explore this space is severely limited due to the significant inherent costs of experimentation and/or...
Join us for a symposium exploring the transformative potential of AI to advance climate tech, hosted by Bakar Climate Labs and Bakar Institute of Digital Materials for the Planet. Objective: This symposium will highlight the appropriate use of AI to enable new climate technologies, showcasing...
Foundation models like GPT-4 have dramatically altered the modern work landscape for many industries reliant on language tasks, but no equivalent model exists yet for scientific applications. Incorporating foundation models into research workflows could enable unprecedented discoveries. However, mainstream foundation models trained on human-scale datasets...
The field of computational chemistry increasingly relies on automated data-driven pipelines designed to facilitate and accelerate the discovery of molecules and materials with tailored properties (i.e., inverse design pipelines). These efforts require extensive infrastructure, integrating quantum chemistry, statistical models, data curation and software to enable...
Modern machine learning has had an outsized impact on many scientific fields, and particle physics is no exception. What is special about particle physics, though, is the vast amount of theoretical knowledge that we already have about many problems in the field, as well as...
Chemical and materials informatics leverages data to correlate a compound’s structure with its properties, enabling the exploration of vast chemical spaces through modeling. Although this approach has the potential to reveal novel materials with desired characteristics, it frequently generates designs that cannot be synthesized, leading...
Frank Noé will explore how AI-powered computational methods are transforming biomolecular research, emphasizing the role of AI in decoding complex biological functions. He will discuss how sophisticated machine learning techniques enhance the analysis of biomolecular dynamics, such as protein folding, offering unprecedented insights into molecular...
After reviewing and updating the theory of energy-conserving Hamiltonian dynamics for optimization and sampling, I'll explain a new application of Energy Conserving Descent (ECD) optimization to precision scientific data analysis in which NN initialization variance has been a bottleneck. Specifically, we choose a particular ECD...
The optimization of catalytic reactions for organic synthesis can be challenging as the interplay between the catalyst structure, reaction conditions, and substrates involved is a complex multidimensional problem. In other words, it is difficult to ascertain the pattern within the noise to offer a complete...
Dark matter is one of the greatest enduring mysteries of fundamental physics. Despite countless direct and indirect searches for dark matter, still, the only evidence we have for it is through its gravitational effects on astrophysical and cosmological scales. In this talk, I will describe...
Transformer-based large language models are making significant strides in various fields, such as natural language processing, biology, chemistry, and computer programming. Here, we show the development and capabilities of Coscientist, an artificial intelligence system that autonomously designs, plans, and performs complex experiments by incorporating large...
The size of chemical space is vast. This makes the application of the first principles of quantum mechanical and advanced statistical mechanics sampling methods to identify binding motifs, conformational equilibria, and reaction pathways extremely challenging, even when considering better physical models, algorithms, or future exascale...
Polycyclic aromatic systems (PASs) present a seemingly insurmountable challenge: vast chemical spaces, complex electronic structures, and elusive aromatic properties. Our mission, should we choose to accept it, is to harness the power of deep learning to decode these molecular mysteries. In this talk, we embark...
Finding particle tracks is a central component of searching for new phenomena, but is very a challenging combinatorial problem. Traditionally, track finding codes assume that tracks must be helical, which simplifies the task but also restricts power to discover new physics which might produce non-helical...
At-scale Human - AI Teams In this talk, Dr. Franziska Bell will share the evolution of enterprise-scale human - AI teams, alongside corresponding examples and share her vision of the next generation of agentic-based AI. Biography Dr. Franziska Bell holds a PhD in theoretical chemistry...