Searcharxiv⌕ Search

arXiv subjects

Ryan P. Adams

Publications and source records attributed to Ryan P. Adams.

At least 19 recordsLinked to original sources

PolyGandr: Steering Generative Models via Spatial Navigation through Latent Space

Generative models support content creation across text, images, videos, and design disciplines, yet their interfaces rely primarily on natural language. While effective for specifying high-level intent, language can be inefficient and imprecise for fine-grained control, iterative refinement, and open-ended exploration. We present PolyGandr, a visual interface for exploring relationships between latent representations and generated outputs, complementing natural-language model steering. PolyGandr formulates generative control as high-dimensional navigation, enabling users to define starting points, steering directions, and step sizes through a unified map-and-mixing-palette interface. Combining global exploration with local refinement supports both broad design-space navigation and fine-grained output manipulation without repeatedly translating visual intent into prompts. We evaluate PolyGandr across three use cases: StyleGAN2 face generation, Stable Diffusion chair design, and rotational kirigami engineering. A formative study indicates that PolyGandr complements existing workflows by supporting real-time editing while simultaneously providing an overview and means to explore the broader design space.

cs.HC↗

Revealing the proton slingshot mechanism in solid acid electrolytes through machine learning molecular dynamics

In solid acid electrolytes CsH$_2$PO$_4$ and CsHSO$_4$, the origin of fast proton conduction has long been debated, attributed either to local proton hopping or to polyanion rotation. However, the precise role of polyanion rotation and its interplay with proton hopping remained unresolved. Nanosecond-scale molecular dynamics simulations, driven by equivariant neural network force fields, reveal a nuanced proton slingshot mechanism: protons are first carried by rotating polyanions, followed by O-H bond reorientation, and the combined motion enables long-range jumps. This mechanism challenges the conventional revolving paddlewheel model, revealing long-range transport arises from significant O-H bond reorientation assisted by only limited polyanion rotation. Despite their structural similarity, CsH$_2$PO$_4$ and CsHSO$_4$ exhibit qualitatively different transport behavior, governed by their distinct proton concentrations. In CsH$_2$PO$_4$, two rotation rates with different activation energies emerge, while CsHSO$_4$ displays a single-rate process with faster productive polyanion dynamics. The higher proton concentration in CsH$_2$PO$_4$ frustrates PO$_4$ rotation, resulting in slower dynamics than for SO$_4$ in CsHSO$_4$. In CsH$_2$PO$_4$, we further identify a unique correlation between oxygen sharing and proton transport, arising from the extra proton per polyanion. Together, our findings establish a unified framework linking polyanion rotation, proton coordination, and long-range transport, and suggest that tuning proton concentration could accelerate rotation and enhance conductivity.

cond-mat.mtrl-sci↗

Coordinate-Free Automatic Differentiation

Automatic differentiation (AD) is a cornerstone of modern machine learning. Since all AD frameworks implicitly fix a basis of Euclidean space to define gradients and Jacobians, they natively process mathematical objects that are arrays of a fixed set of coordinates. This makes it awkward to express objects such as rotations, volume forms, and curvatures that arise, for example, in applications of AI to science. Bespoke workarounds lack reusability. We show that (i) a single, basis-independent derivative operator can represent many geometric differential operations, and (ii) this operator can be expressed entirely in terms of existing AD primitives. This reusable transformation can be composed with JAX batching, compilation, and hardware acceleration. Geometry enters as algebraic data. We experimentally demonstrate its reuse across different manifold operators, its composition through a spacetime inverse-problem pipeline, and its use in optimization over different rotor subgroups.

cs.LG↗

Local Learning Rules for Out-of-Equilibrium Physical Generative Models

We show that the out-of-equilibrium driving protocol of score-based generative models (SGMs) can be learned via local learning rules. The gradient with respect to the parameters of the driving protocol is computed directly from force measurements or from observed system dynamics. As a demonstration, we implement an SGM in a network of driven, nonlinear, overdamped oscillators coupled to a thermal bath. We first apply it to the problem of sampling from a mixture of two Gaussians in 2D. Finally, we train an oscillator network on the MNIST dataset to generate images of handwritten digits 0 and 1.

cs.LG↗

A Single Architecture for Representing Invariance Under Any Space Group

Incorporating known symmetries in data into machine learning models has consistently improved predictive accuracy, robustness, and generalization. However, achieving exact invariance to specific symmetries typically requires designing bespoke architectures for each group, limiting scalability and preventing knowledge transfer across related symmetries. In the case of the space groups, symmetries critical to modeling crystalline solids in materials science and condensed matter physics, this challenge is particularly salient as there are 230 such groups in three dimensions. In this work we present a new approach to such crystallographic symmetries by developing a single machine learning architecture that is capable of adapting its weights automatically to enforce invariance to any input space group. Our approach is based on constructing symmetry-adapted Fourier bases through an explicit characterization of constraints that group operations impose on Fourier coefficients. Encoding these constraints into a neural network layer enables weight sharing across different space groups, allowing the model to leverage structural similarities between groups and overcome data sparsity when limited measurements are available for specific groups. We demonstrate the effectiveness of this approach in achieving competitive performance on material property prediction tasks and performing zero-shot learning to generalize to unseen groups.

cs.LG↗

Space Group Equivariant Crystal Diffusion

Accelerating inverse design of crystalline materials with generative models has significant implications for a range of technologies. Unlike other atomic systems, 3D crystals are invariant to discrete groups of isometries called the space groups. Crucially, these space group symmetries are known to heavily influence materials properties. We propose SGEquiDiff, a crystal generative model which naturally handles space group constraints with space group invariant likelihoods. SGEquiD-iff consists of an SE(3)-invariant, telescoping discrete sampler of crystal lattices; permutation-invariant, transformer-based autoregressive sampling of Wyckoff positions, elements, and numbers of symmetrically unique atoms; and space group equivariant diffusion of atomic coordinates. We show that space group equivariant vector fields automatically live in the tangent spaces of the Wyckoff positions. SGEquiDiff achieves state-of-the-art performance on standard benchmark datasets as assessed by quantitative proxy metrics and quantum mechanical calculations. Our code is available at https://github.com/rees-c/sgequidiff.

cond-mat.mtrl-sci↗

Expressivity of determinantal ansatzes for neural network wave functions

Neural network wave functions have shown promise as a way to achieve high accuracy on the many-body quantum problem. These wave functions most commonly use a determinant or sum of determinants to antisymmetrize many-body orbitals which are described by a neural network. In many cases, the wave function is projected onto a fixed-spin state. Such a treatment is allowed for spin-independent operators; however, it cannot be applied to spin-dependent problems, such as Hamiltonians containing spin-orbit interactions. We show that for spin-independent Hamiltonians, a strict upper bound property is obeyed between a traditional Hartree-Fock like determinant, full spinor wave function, the full determinant wave function, and a generalized spinor wave function. The relationship between a spinor wave function and the full determinant arises because the full determinant wave function is the spinor wave function projected onto a fixed-spin, after which antisymmetry is implicitly restored in the spin-independent case. For spin-dependent Hamiltonians, the full determinant wave function is not applicable, because it is not antisymmetric. Numerical experiments on the H$_3$ molecule and two-dimensional homogeneous electron gas confirm the bounds.

cond-mat.str-el↗

Diagonal Symmetrization of Neural Network Solvers for the Many-Electron Schrödinger Equation

Incorporating group symmetries into neural networks has been a cornerstone of success in many AI-for-science applications. Diagonal groups of isometries, which describe the invariance under a simultaneous movement of multiple objects, arise naturally in many-body quantum problems. Despite their importance, diagonal groups have received relatively little attention, as they lack a natural choice of invariant maps except in special cases. We study different ways of incorporating diagonal invariance in neural network ansätze trained via variational Monte Carlo methods, and consider specifically data augmentation, group averaging and canonicalization. We show that, contrary to standard ML setups, in-training symmetrization destabilizes training and can lead to worse performance. Our theoretical and numerical results indicate that this unexpected behavior may arise from a unique computational-statistical tradeoff not found in standard ML analyses of symmetrization. Meanwhile, we demonstrate that post hoc averaging is less sensitive to such tradeoffs and emerges as a simple, flexible and effective method for improving neural network solvers.

cs.LG↗

Graph Neural Networks Gone Hogwild

Graph neural networks (GNNs) appear to be powerful tools to learn state representations for agents in distributed, decentralized multi-agent systems, but generate catastrophically incorrect predictions when nodes update asynchronously during inference. This failure under asynchrony effectively excludes these architectures from many potential applications where synchrony is difficult or impossible to enforce, e.g., robotic swarms or sensor networks. In this work we identify "implicitly-defined" GNNs as a class of architectures which is provably robust to asynchronous "hogwild" inference, adapting convergence guarantees from work in asynchronous and distributed optimization. We then propose a novel implicitly-defined GNN architecture, which we call an 'energy GNN'. We show that this architecture outperforms other GNNs from this class on a variety of synthetic tasks inspired by multi-agent systems.

cs.LG↗

Real-time design of architectural structures with differentiable mechanics and neural networks

Designing mechanically efficient geometry for architectural structures like shells, towers, and bridges, is an expensive iterative process. Existing techniques for solving such inverse problems rely on traditional optimization methods, which are slow and computationally expensive, limiting iteration speed and design exploration. Neural networks would seem to offer a solution via data-driven amortized optimization, but they often require extensive fine-tuning and cannot ensure that important design criteria, such as mechanical integrity, are met. In this work, we combine neural networks with a differentiable mechanics simulator to develop a model that accelerates the solution of shape approximation problems for architectural structures represented as bar systems. This model explicitly guarantees compliance with mechanical constraints while generating designs that closely match target geometries. We validate our approach in two tasks, the design of masonry shells and cable-net towers. Our model achieves better accuracy and generalization than fully neural alternatives, and comparable accuracy to direct optimization but in real time, enabling fast and reliable design exploration. We further demonstrate its advantages by integrating it into 3D modeling software and fabricating a physical prototype. Our work opens up new opportunities for accelerated mechanical design enhanced by neural networks for the built environment.

cs.CE↗

Generative Marginalization Models

We introduce marginalization models (MAMs), a new family of generative models for high-dimensional discrete data. They offer scalable and flexible generative modeling by explicitly modeling all induced marginal distributions. Marginalization models enable fast approximation of arbitrary marginal probabilities with a single forward pass of the neural network, which overcomes a major limitation of arbitrary marginal inference models, such as any-order autoregressive models. MAMs also address the scalability bottleneck encountered in training any-order generative models for high-dimensional problems under the context of energy-based training, where the goal is to match the learned distribution to a given desired probability (specified by an unnormalized log-probability function such as energy or reward function). We propose scalable methods for learning the marginals, grounded in the concept of "marginalization self-consistency". We demonstrate the effectiveness of the proposed model on a variety of discrete data distributions, including images, text, physical systems, and molecules, for maximum likelihood and energy-based training settings. MAMs achieve orders of magnitude speedup in evaluating the marginal probabilities on both settings. For energy-based training tasks, MAMs enable any-order generative modeling of high-dimensional problems beyond the scale of previous methods. Code is available at https://github.com/PrincetonLIPS/MaM.

cs.LG↗

Designing Mechanical Meta-Materials by Learning Equivariant Flows

Mechanical meta-materials are solids whose geometric structure results in exotic nonlinear behaviors that are not typically achievable via homogeneous materials. We show how to drastically expand the design space of a class of mechanical meta-materials known as cellular solids, by generalizing beyond translational symmetry. This is made possible by transforming a reference geometry according to a divergence free flow that is parameterized by a neural network and equivariant under the relevant symmetry group. We show how to construct flows equivariant to the space groups, despite the fact that these groups are not compact. Coupling this flow with a differentiable nonlinear mechanics simulator allows us to represent a much richer set of cellular solids than was previously possible. These materials can be optimized to exhibit desirable mechanical properties such as negative Poisson's ratios or to match target stress-strain curves. We validate these new designs in simulation and by fabricating real-world prototypes. We find that designs with higher-order symmetries can exhibit a wider range of behaviors.

cs.CE↗

Automated discovery of reprogrammable nonlinear dynamic metamaterials

Harnessing the rich nonlinear dynamics of highly-deformable materials has the potential to unlock the next generation of functional smart materials and devices. However, unlocking such potential requires effective strategies to spatially design optimal material architectures for desired nonlinear dynamic responses such as guiding of nonlinear elastic waves, energy focusing, and cloaking. Here, we introduce an inverse-design framework for the discovery of flexible mechanical metamaterials with a target nonlinear dynamic response. The desired dynamic task is encoded via optimal tuning of the full-scale metamaterial geometry through an inverse-design approach powered by a custom-developed fully-differentiable simulation environment. By deploying such strategy, we design mechanical metamaterials tailored for energy focusing, energy splitting, dynamic protection, and nonlinear motion conversion. Furthermore, we illustrate that our design framework can be expanded to automatically discover reprogrammable architectures capable of switching between different dynamic tasks. For instance, we encode two strongly competing tasks -- energy focusing and dynamic protection -- within a single architecture, utilizing static pre-compression to switch between these behaviors. The discovered designs are physically realized and experimentally tested, demonstrating the robustness of the engineered tasks. All together, our approach opens an untapped avenue towards designer materials with tailored robotic-like reprogrammable functionalities.

physics.app-ph↗

More Stiffness with Less Fiber: End-to-End Fiber Path Optimization for 3D-Printed Composites

In 3D printing, stiff fibers (e.g., carbon fiber) can reinforce thermoplastic polymers with limited stiffness. However, existing commercial digital manufacturing software only provides a few simple fiber layout algorithms, which solely use the geometry of the shape. In this work, we build an automated fiber path planning algorithm that maximizes the stiffness of a 3D print given specified external loads. We formalize this as an optimization problem: an objective function is designed to measure the stiffness of the object while regularizing certain properties of fiber paths (e.g., smoothness). To initialize each fiber path, we use finite element analysis to calculate the stress field on the object and greedily "walk" in the direction of the stress field. We then apply a gradient-based optimization algorithm that uses the adjoint method to calculate the gradient of stiffness with respect to fiber layout. We compare our approach, in both simulation and real-world experiments, to three baselines: (1) concentric fiber rings generated by Eiger, a leading digital manufacturing software package developed by Markforged, (2) greedy extraction on the simulated stress field (i.e., our method without optimization), and (3) the greedy algorithm on a fiber orientation field calculated by smoothing the simulated stress fields. The results show that objects with fiber paths generated by our algorithm achieve greater stiffness while using less fiber than the baselines--our algorithm improves the Pareto frontier of object stiffness as a function of fiber usage. Ablation studies show that the smoothing regularizer is needed for feasible fiber paths and stability of optimization, and multi-resolution optimization helps reduce the running time compared to single-resolution optimization.

cs.CE↗

Gradient-Based Dovetail Joint Shape Optimization for Stiffness

It is common to manufacture an object by decomposing it into parts that can be assembled. This decomposition is often required by size limits of the machine, the complex structure of the shape, etc. To make it possible to easily assemble the final object, it is often desirable to design geometry that enables robust connections between the subcomponents. In this project, we study the task of dovetail-joint shape optimization for stiffness using gradient-based optimization. This optimization requires a differentiable simulator that is capable of modeling the contact between the two parts of a joint, making it possible to reason about the gradient of the stiffness with respect to shape parameters. Our simulation approach uses a penalty method that alternates between optimizing each side of the joint, using the adjoint method to compute gradients. We test our method by optimizing the joint shapes in three different joint shape spaces, and evaluate optimized joint shapes in both simulation and real-world tests. The experiments show that optimized joint shapes achieve higher stiffness, both synthetically and in real-world tests.

cs.CE↗

JAX FDM: A differentiable solver for inverse form-finding

We introduce JAX FDM, a differentiable solver to design mechanically efficient shapes for 3D structures conditioned on target architectural, fabrication and structural properties. Examples of such structures are domes, cable nets and towers. JAX FDM solves these inverse form-finding problems by combining the force density method, differentiable sparsity and gradient-based optimization. Our solver can be paired with other libraries in the JAX ecosystem to facilitate the integration of form-finding simulations with neural networks. We showcase the features of JAX FDM with two design examples. JAX FDM is available as an open-source library at https://github.com/arpastrana/jax_fdm.

cs.CE↗

A rapid and automated computational approach to the design of multistable soft actuators

We develop an automated computational modeling framework for rapid gradient-based design of multistable soft mechanical structures composed of non-identical bistable unit cells with appropriate geometric parameterization. This framework includes a custom isogeometric analysis-based continuum mechanics solver that is robust and end-to-end differentiable, which enables geometric and material optimization to achieve a desired multistability pattern. We apply this numerical modeling approach in two dimensions to design a variety of multistable structures, accounting for various geometric and material constraints. Our framework demonstrates consistent agreement with experimental results, and robust performance in designing for multistability, which facilities soft actuator design with high precision and reliability.

math.NA↗

Representing and Learning Functions Invariant Under Crystallographic Groups

Crystallographic groups describe the symmetries of crystals and other repetitive structures encountered in nature and the sciences. These groups include the wallpaper and space groups. We derive linear and nonlinear representations of functions that are (1) smooth and (2) invariant under such a group. The linear representation generalizes the Fourier basis to crystallographically invariant basis functions. We show that such a basis exists for each crystallographic group, that it is orthonormal in the relevant $L_2$ space, and recover the standard Fourier basis as a special case for pure shift groups. The nonlinear representation embeds the orbit space of the group into a finite-dimensional Euclidean space. We show that such an embedding exists for every crystallographic group, and that it factors functions through a generalization of a manifold called an orbifold. We describe algorithms that, given a standardized description of the group, compute the Fourier basis and an embedding map. As examples, we construct crystallographically invariant neural networks, kernel machines, and Gaussian processes.

stat.ML↗