arXiv subjects

Greg van Anders

Publications and source records attributed to Greg van Anders.

At least 19 recordsLinked to original sources

arXiv2026

Causal Atlases from Entropic Inference: Bayesian Networks beyond Optimal DAGs

Data-driven causal relationship identification is pertinent to advancing understanding of complex systems both within and beyond science. Bayesian networks offer a probabilistic method for modelling generic causal relationships via directed acyclic graphs (DAGs). However, typical techniques for constructing Bayesian networks rely on optimization, which can be ill-suited for learning causal relationships because the underlying data may admit multiple chains of causation. More data-faithful representations of causal relationships would provide frameworks for constructing multiple causal maps that are consistent with the variability that is inherent in underlying data. Here, we show that entropy-based inference generates atlases of plausible causal relationships that are consistent with underlying data. On simulated noisy data of 2- and 20-node linear structural equation models, we sample a maximum-entropy ensemble of graphs that allow us to quantify the inherent structural ambiguity in underlying causal relationships. Our method shows that "optimized" DAGs can contain causal artifacts are not consistent across equivalently accurate topologies.

cs.LG↗

arXiv2026

Entropic Auto-Encoding via Implicit Free-Energy Minimization

Despite their ubiquity, variational autoencoders (VAEs) inherently suffer from posterior collapse, a failure mode in which latent variables are effectively ignored. This failure arises because explicit prior imposition drives optimization toward loss landscape regions corresponding to uninformative latent representations. Here, we introduce Entropic Autoencoders (EAEs), a framework in which reconstruction loss is the only explicit objective, and entropy generates the latent variables' prior implicitly through a free energy-minimizing ensemble of encoders. This ensemble biases learning toward high-volume regions of near-optimal solutions, while decoder updates direct the search trajectories toward informative latent representations. We demonstrate that EAEs mitigate posterior collapse by learning non-Gaussian, multimodal latent distributions that yield diverse, data-consistent generations and preserve different forms of underlying structure in the data. As a proof-of-concept, we show that an EAE captures a superposition of the known low-dimensional dynamics of a reaction-diffusion process. Then, we show that an EAE identifies implicit categorical distinctions in MNIST latent representations, and displays a hierarchical understanding of facial structure on the CelebA dataset, from an "all-human" face to individual-dependent features.

cs.LG↗

arXiv2025

Molecular Dynamics Simulations of Bubble Nucleation in a Liquid-Noble Scintillator

The Scintillating Bubble Chamber collaboration is searching for Weakly Interacting Massive Particles using a novel bubble chamber with intended thresholds as low as 100eV. Existing molecular dynamics simulations of bubble formation in bubble chambers were conducted with non-scintillating target materials and therefore do not account for the energy transfer to photons or time-delayed releases that occur in atomic de-excitation. In this study, we use the HOOMD-blue molecular dynamics framework to simulate bubble formation in liquid argon, including photon creation, ionization, and direct nuclear recoils. A multi-stage bubble growth process similar to that reported in the literature was observed. When comparing simulated thresholds with and without scintillation effects, we found that scintillation raises the average energy required to form a bubble by a factor of 2.16. This is larger than the fraction of energy lost to photon creation, and demonstrates that energy stored in excited molecular states with lifetimes longer than the rapid growth phase of nucleation (~250 ps) does not contribute significantly to bubble formation. This conclusion was further supported by simulations showing increased bubble nucleation thresholds when the excited molecular state lifetimes were increased, even under identical thermodynamic conditions.

hep-ex↗

arXiv2025

Filters reveal emergent structure in computational morphogenesis

Revolutionary advances in both manufacturing and computational morphogenesis raise critical questions about design sensitivity. Sensitivity questions are especially critical in contexts, such as topology optimization, that yield structures with emergent morphology. However, analyzing emergent structures via conventional, perturbative techniques can mask larger-scale vulnerabilities that could manifest in essential components. Risks that fail to appear in perturbative sensitivity analyses will only continue to proliferate as topology optimization-driven manufacturing penetrates more deeply into engineering design and consumer products. Here, we introduce Laplace-transform based computational filters that supplement computational morphogenesis with a set of nonperturbative sensitivity analyses. We demonstrate how this approach identifies important elements of a structure even in the absence of knowledge of the ultimate, optimal structure itself. We leverage techniques from molecular dynamics and implement these methods in open-source codes, demonstrating their application to compliance minimization problems in both 2D and 3D. Our implementation extends straightforwardly to topology optimization for other problems and benefits from the strong scaling properties observed in conventional molecular simulation.

cs.CE↗

arXiv2024

Simmering: Sufficient is better than optimal for training neural networks

The broad range of neural network training techniques that invoke optimization but rely on ad hoc modification for validity suggests that optimization-based training is misguided. Shortcomings of optimization-based training are brought to particularly strong relief by the problem of overfitting, where naive optimization produces spurious outcomes. The broad success of neural networks for modelling physical processes has prompted advances that are based on inverting the direction of investigation and treating neural networks as if they were physical systems in their own right. These successes raise the question of whether broader, physical perspectives could motivate the construction of improved training algorithms. Here, we introduce simmering, a physics-based method that trains neural networks to generate weights and biases that are merely ``good enough'', but which, paradoxically, outperforms leading optimization-based approaches. Using classification and regression examples we show that simmering corrects neural networks that are overfit by Adam, and show that simmering avoids overfitting if deployed from the outset. Our results question optimization as a paradigm for neural network training, and leverage information-geometric arguments to point to the existence of classes of sufficient training algorithms that do not take optimization as their starting point.

cs.LG↗

arXiv2024

Transforming Design Spaces Using Pareto-Laplace Filters

Optimization is a critical tool for addressing a broad range of human and technical problems. However, the paradox of advanced optimization techniques is that they have maximum utility for problems in which the relationship between the structure of the problem and the ultimate solution is the most obscure. The existence of solution with limited insight contrasts with techniques that have been developed for a broad range of engineering problems where integral transform techniques yield solutions and insight in tandem. Here, we present a ``Pareto-Laplace'' integral transform framework that can be applied to problems typically studied via optimization. We show that the framework admits related geometric, statistical, and physical representations that provide new forms of insight into relationships between objectives and outcomes. We argue that some known approaches are special cases of this framework, and point to a broad range of problems for further application.

cs.CE↗

arXiv2023

SLO/GO Degradation-Loss Sensitivity in Climate-Human System Coupling

The potential of extreme environmental change driven by a destabilized climate system is an alarming prospect for humanity. But the intricate, subtle ways Earth's climate couples to social and economic systems raise the question of when more incremental climate change signals the need for alarm. Questions about incremental sensitivity are particularly crucial for human systems that are organized by optimization. Optimization is most valuable in resolving complex interactions among multiple factors, however, those interactions can obscure coupling to underlying drivers such as environmental degradation. Here, using Multi-Objective Land Allocation as an example, we show that model features that are common across non-convex optimization problems drive hypersensitivities in climate-induced degradation--loss response. We show that catastrophic losses in human systems can occur well before catastrophic climate collapse. We find punctuated insensitive/hypersensitive degradation--loss response, which we trace to the contrasting effects of environmental degradation on subleading, local versus global optima (SLO/GO). We argue that the SLO/GO response we identify in land-allocation problems traces to features that are common across non-convex optimization problems more broadly. Given the broad range of human systems that rely on non-convex optimization, our results therefore suggest that substantial social and economic risks could be lurking in a broad range in human systems that are coupled to the environment, even in the absence of catastrophic changes to the environment itself.

cs.CE↗

arXiv2023

No Free Lunch for Avoiding Clustering Vulnerabilities in Distributed Systems

Emergent design failures are ubiquitous in complex systems, and often arise when system elements cluster. Approaches to systematically reduce clustering could improve a design's resilience, but reducing clustering is difficult if it is driven by collective interactions among design elements. Here, we use techniques from statistical physics to identify mechanisms by which spatial clusters of design elements emerge in complex systems modelled by heterogeneous networks. We find that, in addition to naive, attraction-driven clustering, heterogeneous networks can exhibit emergent, repulsion-driven clustering. We draw quantitative connections between our results on a model system in naval engineering to entropy-driven phenomena in nanoscale self-assembly, and give a general argument that the clustering phenomena we observe should arise in many distributed systems. We identify circumstances under which generic design problems will exhibit trade-offs between clustering and uncertainty in design objectives, and we present a framework to identify and quantify trade-offs to manage clustering vulnerabilities.

physics.soc-ph↗

arXiv2023

Flashpoints Signal Hidden Inherent Instabilities in Land-Use Planning

Land-use decision-making processes have a long history of producing globally pervasive systemic equity and sustainability concerns. Quantitative, optimization-based planning approaches, e.g. Multi-Objective Land Allocation (MOLA), seemingly open the possibility to improve objectivity and transparency by explicitly evaluating planning priorities by the type, amount, and location of land uses. Here, we show that optimization-based planning approaches with generic planning criteria generate a series of unstable "flashpoints" whereby tiny changes in planning priorities produce large-scale changes in the amount of land use by type. We give quantitative arguments that the flashpoints we uncover in MOLA models are examples of a more general family of instabilities that occur whenever planning accounts for factors that coordinate use on- and between-sites, regardless of whether these planning factors are formulated explicitly or implicitly. We show that instabilities lead to regions of ambiguity in land-use type that we term "gray areas". By directly mapping gray areas between flashpoints, we show that quantitative methods retain utility by reducing combinatorially large spaces of possible land-use patterns to a small, characteristic set that can engage stakeholders to arrive at more efficient and just outcomes.

cs.AI↗

arXiv2021

Shape and Interaction Decoupling for Colloidal Pre-Assembly

Creating materials with structure that is independently controllable at a range of scales requires breaking naturally occurring hierarchies. Breaking these hierarchies can be achieved via the decoupling of building block attributes from structure during assembly. Here, we demonstrate both geometric and interaction decoupling in pre-assembled colloidal structures of cube-like particles with rounded edges. Through computer simulations and experiments, we show that compressing a small number of such cubes in spherical confinement results in clusters with highly reproducible structures that can be used as mesoscale building blocks to form the next level of structural hierarchy. These clusters demonstrate geometric decoupling between particle shape and cluster structure; namely, for clusters of up to nine particles, the colloidal superballs pack consistently like spheres, despite the presence of shape anisotropy and facets in the cubic-like particles. We confirm that cluster structure is also decoupled from inter-particle interaction, showing that the same structures arise from the spherical confinement of both non-magnetic and magnetic colloidal cubes with strong dipolar interactions. To highlight the potential of these superball clusters for hierarchical assembly, we demonstrate, using computer simulations, that clusters of six to nine particles can self-assemble into high-order structures that differ from those of similarly shaped particles without pre-assembly. These results demonstrate decoupling for anisotropic building blocks that can be further exploited for hierarchical materials development.

cond-mat.soft↗

arXiv2021

Avoidance, Adjacency, and Association in Distributed Systems Design

Patterns of avoidance, adjacency, and association in complex systems design emerge from the system's underlying logical architecture (functional relationships among components) and physical architecture (component physical properties and spatial location). Understanding the physical--logical architecture interplay that gives rise to patterns of arrangement requires a quantitative approach that bridges both descriptions. Here, we show that statistical physics reveals patterns of avoidance, adjacency, and association across sets of complex, distributed system design solutions. Using an example arrangement problem and tensor network methods, we identify several phenomena in complex systems design, including placement symmetry breaking, propagating correlation, and emergent localization. Our approach generalizes straightforwardly to a broad range of complex systems design settings where it can provide a platform for investigating basic design phenomena.

physics.soc-ph↗

arXiv2020

Robust Design from Systems Physics

A crucial challenge in engineering modern, integrated systems is to produce robust designs. Ensuring robust design is difficult because subsystem couplings produce unpredictable response to changes in whole system specifications. Here, we show that the response of design elements to whole-system specification changes can be characterized, as materials are, using strong/weak and brittle/ductile dichotomies. We find these dichotomies emerge from a mesoscale treatment of early stage design problems that we cast in terms of stress-strain relationships. Compared with other state-of-the-art techniques, we propose a two-factor robustness metric that is computable for sets of design solutions. We illustrate the use of this approach with examples of arrangement problems drawn from naval engineering, however our approach is immediately applicable to a broad range of problems in integrated systems design.

physics.soc-ph↗

arXiv2019

Topological Order in Densely Packed Anisotropic Colloids

The existence of topological order is frequently associated with strongly coupled quantum matter. Here, we demonstrate the existence of topological phases in classical systems of densely packed, hard, anisotropic polyhedrally shaped colloidal particles. We show that previously reported transitions in dense packings lead to the existence of topologically ordered thermodynamic phases, which we show are stable away from the dense packing limit. Our work expands the library of known topological phases, whose experimental realization could provide new means for constructing plasmonic materials that are robust in the presence of fluctuations.

cond-mat.soft↗

arXiv2019

FCC-to-BCC phase transitions in convex and concave hard particle systems

Particle shape plays an important role in the phase behavior of colloidal self-assembly. Recent progress in particle synthesis has made particles of polyhedral shapes and dimpled spherical shapes available. Here using computer simulations of hard particle models, we study face-centered cubic to body-centered cubic (FCC-to-BCC) phase transitions in a convex 432 polyhedral shape family and a concave dimpled sphere family. Particles in both families have four-, three-, and two-fold rotational symmetries. Via free energy calculations we find the FCC-to-BCC transitions in both families are first order. As a previous work reports the FCC-to-BCC phase transition is first order in a convex 332 family of hard polyhedra, our work provides additional insight into the FCC-to-BCC transition and how the convexity or concavity of particle shape affects phase transition pathways.

cond-mat.soft↗

arXiv2018

Statistical Physics of Design

A key challenge in complex design problems that permeate science and engineering is the need to balance design objectives for specific design elements or subsystems with global system objectives. Global objectives give rise to competing design pressures, whose effects can be difficult to trace in subsystem design. Here, using examples from arrangement problems, we show that the systems-level application of statistical physics principles, which we term "systems physics", provides a detailed characterization of subsystem design in terms of the concepts of stress and strain from materials physics. We analyze instances of routing problems in naval architectures, and show that systems physics provides a direct means of classifying architecture types, and quantifying trade-offs between subsystem- and overall performance. Our approach generalizes straightforwardly to design problems in a wide range of other disciplines that require concrete understanding of how the pressure to meet overall design objectives drives the outcomes for component subsystems.

physics.soc-ph↗

arXiv2018

Pressure-Tunable Photonic Band Gaps in an Entropic Colloidal Crystal

Materials adopting the diamond structure possess useful properties in atomic and colloidal systems, and are a popular target for synthesis in colloids where a photonic band gap is possible. The desirable photonic properties of the diamond structure pose an interesting opportunity for reconfigurable matter: can we create a colloidal crystal able to switch reversibly to and from the diamond structure? Drawing inspiration from high-pressure transitions of diamond-forming atomic systems, we design a system of polyhedrally-shaped particles that transitions from diamond to a tetragonal diamond derivative upon a small pressure change. The transition can alternatively be triggered by changing the shape of the particle in-situ. We propose that the transition provides a reversible reconfiguration process for a potential new colloidal material, and draw parallels between this transition and phase behavior of the atomic transitions from which we take inspiration.

cond-mat.mtrl-sci↗

arXiv2018

Mapping Disorder in Entropically Ordered Crystals

Systems of hard shapes crystallize due to entropy. How is entropy distributed among translational and rotational microscopic contributions? We answer this question by decomposing thermal fluctuation of crystals of hard hexagons into collective modes, a generalization and quantification of the Onsager picture of hard rod liquid crystals. We show that at densities both near densest packing and near the solid-hexatic melting transition, solids of hard regular hexagons hold most of their entropy in translational degrees of freedom.

cond-mat.soft↗

arXiv2018

Predicting colloidal crystals from shapes via inverse design and machine learning

A fundamental challenge in materials design is linking building block attributes to crystal structure. Addressing this challenge is particularly difficult for systems that exhibit emergent order, such as entropy-stabilized colloidal crystals. We combine recently developed techniques in inverse design with machine learning to construct a model that correctly classifies the crystals of more than ten thousand polyhedral shapes into 13 different structures with a predictive accuracy of 96% using only two geometric shape measures. With three measures, 98% accuracy is achieved. We test our model on previously reported colloidal crystal structures for 71 symmetric polyhedra and obtain 92% accuracy. Our findings (1) demonstrate that entropic colloidal crystals are controlled by surprisingly few parameters, (2) provide a quantitative model to predict these crystals solely from the geometry of their building blocks, and (3) suggest a prediction paradigm that easily generalizes to other self-assembled materials.

cond-mat.mtrl-sci↗