arXiv research

A new multi-hop FiLM approach improves visual reasoning tasks.

Efficient neural models for complex multi-hop reasoning tasks.

New AI algorithm improves multi-layer optical film design efficiency.

In this paper we first review the covering space method with constrained BV functions for solving the classical Plateau's problem. Next, we carefully analyze some interesting examples of soap films compatible with the covering space method: in particular, the case of a soap film only partially wetting a space curve, a …

Solves the film scheduling and staggered showtimes problem for movie theaters.

A novel beam training scheme optimizes multi-hop THz communications with up to 75% performance gain.

Method integrates logical rules into neural multi-hop reasoning for drug repurposing.

In this paper we provide the first examples of non-flat soap films proven to span tetrahedra. These are members of a continuous two parameter family of soap films with tetrahedral boundaries. Of particular interest is a two parameter subfamily where each spanning soap film has the property that two minimal surfaces mee…

Soap films collapse only if their bulk has negative pressure, forming convex shapes.

Derives equilibrium law for Plateau borders in wet soap films and foams.

New method uses multi-hop assortativity to classify network functionalities.

We introduce a general-purpose conditioning method for neural networks called FiLM: Feature-wise Linear Modulation. FiLM layers influence neural network computation via a simple, feature-wise affine transformation based on conditioning information. We show that FiLM layers are highly effective for visual reasoning - an…

Extends symmetry and rigidity to surfaces with soap film-like singularities.

We prove existence and a.e. regularity of an area minimizing soap film with a bound on energy spanning a given Jordan curve in R^3. The energy of a film is defined to be the sum of its surface area and the length of its singular branched set. The class of surfaces over which area is minimized includes images of disks, …

A soap film is actually a thin solid fluid bounded by two surfaces of opposite orientation. It is natural to model the film using one polyhedron for each side. Two problems are to get the polyhedra for both sides to be in the same place without canceling each other out and to model triple junctions without introducing …

Develops a geometric framework for soap film instabilities.

GNN-FiLM uses feature-wise linear modulation to improve graph neural networks.

MAGNA improves graph neural networks by incorporating multi-hop context information.

What are the possible shapes of various things and why? For instance, when a closed wire or a frame is dipped into a soap solution and is raised up from the solution, the surface spanning the wire is a soap film. What are the possible shapes of soap films and why? Or, for instance, why is DNA like a double spiral stair…

Smoothness of collapsed regions in soap films is proven, indicating wetted singularities.

Classifies soap film surfaces with vertical potentials.

For a given boundary set consisting of arcs and vertices, with two or more arcs meeting at each vertex, we treat the problem of estimating the area density of a soap film-like surface spanning the boundary.

Proposes a framework for private data augmentation in federated learning.

A deep learning algorithm designs low-cost SPP films.

FiLM improves deep learning for long-term time series forecasting.

We develop a theory of axisymmetric surfaces minimizing a combination of surface tension and nematic elastic energies which may be suitable for describing simple film and bubble shapes. As a function of the elastic constant and the applied tension on the bubbles, we find the analogues of the unduloid, sphere, and nodoi…

Plateau's soap film problem is to find a surface of least area spanning a given boundary. We begin with a compact orientable $(n-2)$-dimensional submanifold $M$ of $\R^n$. If $M$ is connected, we say a compact set $X$ "spans" $M$ if $X$ intersects every Jordan curve whose linking number with $M$ is 1. Picture a soap fi…

Automated malaria diagnosis from field slides achieves accurate results.

Improved continual learning method using variational inference and FiLM layers.

Let $M$ be an $n$-dimensional complete simply connected Riemannian manifold with sectional curvature bounded above by a nonpositive constant $-κ^2$. Using the cone total curvature $TC(Γ)$ of a graph $Γ$ which was introduced by Gulliver and Yamada Math. Z. 2006, we prove that the density at any point of a soap film-like…

Motivated by the study of the equilibrium equations for a soap film hanging from a wire frame, we prove a compactness theorem for surfaces with asymptotically vanishing mean curvature and fixed or converging boundaries. In particular, we obtain sufficient geometric conditions for the minimal surfaces spanned by a given…

New method finds all thin film structures from reflectometry data.

Superhighway bypasses data sparsity in cross-domain CF.

New neural KB representation speeds up reasoning with large symbolic knowledge bases.

In mathematics, the classical Plateau problem consists of finding the surface of least area that spans a given rigid boundary curve. A physical realization of the problem is obtained by dipping a stiff wire frame of some given shape in soapy water and then removing it; the shape of the spanning soap film is a solution …

Graph neural networks improve volatility forecasting by capturing spillover effects.

Knowledge base (KB) completion adds new facts to a KB by making inferences from existing facts, for example by inferring with high likelihood nationality(X,Y) from bornIn(X,Y). Most previous methods infer simple one-hop relational synonyms like this, or use as evidence a multi-hop relational path treated as an atomic f…

This paper proves that classical minimal surfaces of arbitrary topological type with total boundary curvature at most 4πmust be smoothly embedded. Related results are proved for varifolds and for soap film surfaces.

FinReflectKG benchmarks financial QA by linking relevant context from a financial KG, improving model performance and efficiency.

Langmuir-Blodgett films (LB-films) consist from few LB-monolayers which are high structured nanomaterials that are very promising materials for applications. We use a geometrical approach to describe structurization into LB-monolayers. Consequently, we develop on the 1-jet space J^1([0,\infty),R^2) the single-time Lagr…

Machine learning speeds up FLIM analysis in biomedical research.

This study compares GNNs and GA-MLPs, finding GA-MLPs can distinguish graphs but not count walks.

New task AQA tackles acoustic reasoning from sound scenes.

TFiLM expands convolutional models' receptive field with minimal overhead.

We derive a new model for pre-strained thin films, which consists of minimizing a biharmonic energy of deformations $v\in W^{2,2}$ satisfying the Monge-Ampère constraint $\det\nabla^2v = f$. We further discuss multiplicity properties of the minimizers of this model, in some special cases.

Embedding models for entities and relations are extremely useful for recovering missing facts in a knowledge base. Intuitively, a relation can be modeled by a matrix mapping entity vectors. However, relations reside on low dimension sub-manifolds in the parameter space of arbitrary matrices---for one reason, compositio…

TSDS framework reduces edge LLM agent compute by 43%-73% while maintaining safety and reliability.

Multi-hop inference is necessary for machine learning systems to successfully solve tasks such as Recognising Textual Entailment and Machine Reading. In this work, we demonstrate the effectiveness of adaptive computation for learning the number of inference steps required for examples of different complexity and that l…