arXiv subjects

Marjan Famili

Publications and source records attributed to Marjan Famili.

5 recordsLinked to original sources

arXiv2026

Reliability of Probabilistic Emulation of Physical Systems

Two dominant approaches have emerged for generating probabilistic forecasts of physical systems: generative models, such as diffusion or flow matching; and ensembles of deterministic models with stochasticity injected, trained using the continuous ranked probability score (CRPS) loss. While both approaches have demonstrated strong predictive accuracy, the reliability of their uncertainties has not been systematically assessed. We address this gap by developing a framework to evaluate both approaches across diverse 2D spatiotemporal physical systems, under matched model size and computational budget. We assess the reliability of probabilistic emulation by inspecting the empirical coverage of predictive intervals, while also considering accuracy and computational efficiency metrics. CRPS-trained ensembles typically achieve more reliable uncertainties on both single-step prediction and autoregressive rollouts, demonstrating better coverage than the standard alternative of training generative models in a latent space. Moreover, the CRPS approach offers significantly faster inference. When generative models are trained in ambient rather than a compressed latent space, which is often infeasible for high-dimensional problems, they exhibit comparable coverage to CRPS-trained ensembles, though with substantially larger inference latency. In contrast, when CRPS-trained ensembles are trained in latent space they do not show a marked degradation in coverage with respect to ambient space. Both generative models and CRPS-trained ensembles demonstrate good predictive accuracy. To facilitate future research and application, we release AutoCast, a modular framework implementing both generative models and CRPS-trained ensembles, alongside AutoSim, a flexible dataset generation package for rapid prototyping.

cs.LG↗

arXiv2025

PERC: a suite of software tools for the curation of cryoEM data with application to simulation, modelling and machine learning

Ease of access to data, tools and models expedites scientific research. In structural biology there are now numerous open repositories of experimental and simulated datasets. Being able to easily access and utilise these is crucial for allowing researchers to make optimal use of their research effort. The tools presented here are useful for collating existing public cryoEM datasets and/or creating new synthetic cryoEM datasets to aid the development of novel data processing and interpretation algorithms. In recent years, structural biology has seen the development of a multitude of machine-learning based algorithms for aiding numerous steps in the processing and reconstruction of experimental datasets and the use of these approaches has become widespread. Developing such techniques in structural biology requires access to large datasets which can be cumbersome to curate and unwieldy to make use of. In this paper we present a suite of Python software packages which we collectively refer to as PERC (profet, EMPIARreader and CAKED). These are designed to reduce the burden which data curation places upon structural biology research. The protein structure fetcher (profet) package allows users to conveniently download and cleave sequences or structures from the Protein Data Bank or Alphafold databases. EMPIARreader allows lazy loading of Electron Microscopy Public Image Archive datasets in a machine-learning compatible structure. The Class Aggregator for Key Electron-microscopy Data (CAKED) package is designed to seamlessly facilitate the training of machine learning models on electron microscopy data, including electron-cryo-microscopy-specific data augmentation and labelling. These packages may be utilised independently or as building blocks in workflows. All are available in open source repositories and designed to be easily extensible to facilitate more advanced workflows if required.

cs.LG↗

arXiv2025

Affinity-VAE: incorporating prior knowledge in representation learning from scientific images

Learning compact and interpretable representations of data is a critical challenge in scientific image analysis. Here, we introduce Affinity-VAE, a generative model that enables us to impose our scientific intuition about the similarity of instances in the dataset on the learned representation during training. We demonstrate the utility of the approach in the scientific domain of cryo-electron tomography (cryo-ET) where a significant current challenge is to identify similar molecules within a noisy and low contrast tomographic image volume. This task is distinct from classification in that, at inference time, it is unknown whether an instance is part of the training set or not. We trained affinity-VAE using prior knowledge of protein structure to inform the latent space. Our model is able to create rotationally-invariant, morphologically homogeneous clusters in the latent representation, with improved cluster separation compared to other approaches. It achieves competitive performance on protein classification with the added benefit of disentangling object pose, structural similarity and an interpretable latent representation. In the context of cryo-ET data, affinity-VAE captures the orientation of identified proteins in 3D which can be used as a prior for subsequent scientific experiments. Extracting physical principles from a trained network is of significant importance in scientific imaging where a ground truth training set is not always feasible.

cs.CV↗

arXiv2022

Ab initio electronic stationary states for nuclear projectiles in solids

The process by which a nuclear projectile is decelerated by the electrons of the condensed matter it traverses is currently being studied by following the explicit dynamics of projectile and electrons from first principles in a simulation box with a sample of the host matter in periodic boundary conditions. The approach has been quite successful for diverse systems even in the strong-coupling regime of maximal dissipation. This technique is here revisited for periodic solids in the light of the Floquet theory of stopping, a time-periodic scattering framework characterizing the stationary dynamicalsolutions for a constant velocity projectile in an infinite solid. The effect of proton projectiles in diamond is studied under that light, using time-dependent density-functional theory in real time. The Floquet quasi-energy conserving stationary scattering regime, characterized by time-periodic properties such as particle density and the time derivative of energy, is obtained for a converged system size of one thousand atoms. The validity of the customary calculation of electronic stopping power from the average slope of the density-functional total energy is discussed. Quasi-energy conservation, as well as the implied fundamental approximations, are critically reviewed.

cond-mat.other↗

arXiv2021

Local orbital formulation of the Floquet theory of projectile electronic stopping

A recently proposed theoretical framework for the description of electronic quantum friction for constant-velocity nuclear projectiles traversing periodic crystals is here implemented using a local basis representation. The theory requires a change of reference frame to that of the projectile, and a basis set transformation for the target basis functions to a gliding basis is presented, which is time-periodic but does not displace in space with respect to the projectile, allowing a local-basis Floquet impurity-scattering formalism to be used. It is illustrated for a one-dimensional single-band tight-binding model, as the simplest paradigmatic example, displaying the qualitative behaviour of the formalism. The time-dependent non-orthogonality of the gliding basis requires care in the proper (simplest) definition of a local projectile perturbation. The Fermi level is tilted with a slope given by the projectile velocity, which complicates integration over occupied states. It is solved by a recurrent application of the Lippmann-Schwinger equation, in analogy with previous non-equilibrium treatment of electron ballistic transport. Aiming towards a first-principles mean-field-like implementation, the final result is the time-periodic particle density in the region around the projectile, describing the stroboscopically stationary perturbation cloud around the projectile, out of which other quantities can be obtained, such as the electronic stopping power.

cond-mat.other↗