New metrics for Anosov representations defined from Thurston's asymmetric metrics.
problem Defining metrics for Anosov representations.
method Generalizing Thurston's asymmetric metric to Anosov representations.
result Provides a (possibly asymmetric) Finsler distance in some cases.
Novel representer theorem for metric and preference learning in RKHSs.
problem Metric and preference learning problems in Hilbert spaces.
method Regularization with respect to task structure norm, RKHS representation, and novel algorithm.
result Significant performance improvement over baseline methods in real-world rank inference benchmarks.
Study Eisenstein metrics on modular group representations.
problem Harmonic metrics on automorphic vector bundles.
method Eisenstein series construction for metrics.
result Residue of Eisenstein metrics is a harmonic metric.
New metric for disentangling multivariate representations, accounting for more complex entanglements.
problem Current disentanglement metrics fail to detect entanglements involving more than two variables.
method Partial Information Decomposition framework to analyze information sharing and propose a new disentanglement metric.
result The proposed metric correctly identifies entanglements in high-dimensional spaces.
Study pressure metrics for cusped Hitchin representations.
problem Characterize cusped Hitchin representations of Fuchsian groups.
method Develop pressure metrics associated to fundamental weights and roots.
result New pressure metrics for Hilbert length when d=3. We analyze disentangled representations under a causal generative process, proposing new metrics and datasets.
problem Addressing fairness and interpretability through disentangled representations with a causal perspective.
method Work under a causal generative process, proposing new metrics and datasets to study disentanglement.
result Proposed metrics capture the desiderata of disentangled causal process.
We relate Ambrosio-Kirchheim metric currents to Alberti representations and Weaver derivations. In particular, given a metric current T, we show that if the module X(∥T∥) of Weaver derivations is finitely generated, then T can be represented in terms of derivations; this extends previous results of Wi…
Using the thermodynamics formalism, we introduce a notion of intersection for projective Anosov representations, show analyticity results for the intersection and the entropy, and rigidity results for the intersection. We use the renormalized intersection to produce a Out(Γ)-invariant Riemannian metric on the smooth …
New metric captures individual neuron tuning across neural networks.
problem Need a metric that respects individual neuron tuning across different neural networks.
method Derived a 'soft' permutation-based metric using optimal transport theory.
result Metric avoids counter-intuitive outcomes and captures geometric insights.
We investigate orthogonal representations of compact Lie groups from the point of view of their quotient spaces, considered as metric spaces. We study metric spaces which are simultaneously quotients of different representations and investigate properties of the corresponding representations. We obtain some structural …
Defines metrics to compare neural network representations.
problem Comparing neural network representations across different architectures and tasks.
method Developed a family of metric spaces and modified existing measures to quantify representational dissimilarity.
result Identified relationships between neural representations and anatomical features.
A new geometric metric identifies true data changes from parametrization artifacts in high-dimensional representations.
problem Quantifying representation drift in high-dimensional data using Euclidean or cosine distances can misattribute changes due to arbitrary parametrizations.
method Introducing the Fubini Study metric to identify representations that differ only by gauge transformations.
result The Fubini Study metric isolates intrinsic evolution by remaining invariant under gauge-induced fluctuations, providing a diagnostic for meaningful structural changes.
State representation learning aims at learning compact representations from raw observations in robotics and control applications. Approaches used for this objective are auto-encoders, learning forward models, inverse dynamics or learning using generic priors on the state characteristics. However, the diversity in appl…
We make two theoretical contributions to disentanglement learning by (a) defining precise semantics of disentangled representations, and (b) establishing robust metrics for evaluation. First, we characterize the concept "disentangled representations" used in supervised and unsupervised methods along three dimensions-in…
In this paper we give a spinorial representation of submanifolds of any dimension and codimension into Lie groups equipped with left invariant metrics. As applications, we get a spinorial proof of the Fundamental Theorem for submanifolds into Lie groups, we recover previously known representations of submanifolds in $\…
New measures quantify diversity of latent representations using metric space magnitude.
problem Evaluating the diversity of latent representations in machine learning models.
method Developed magnitude-based measures for latent representations, stable under data perturbations.
result Demonstrated superior performance across various domains and tasks.
As a highlighting research topic in the multimedia area, cross-media retrieval aims to capture the complex correlations among multiple media types. Learning better shared representation and distance metric for multimedia data is important to boost the cross-media retrieval. Motivated by the strong ability of deep neura…
This paper evaluates fairness in deep metric learning and proposes a method to reduce subgroup performance gaps.
problem The negative impact of deep metric learning representations on minority subgroup performance in downstream tasks.
method Definition of fairness in DML through inter-class, intra-class, and uniformity properties; finDML benchmark; Partial Attribute De-correlation (PARADE) method.
result Bias in DML representations propagates to downstream tasks, even with balanced training data.
To evaluate disentangled representations several metrics have been proposed. However, theoretical guarantees for conventional metrics of disentanglement are missing. Moreover, conventional metrics do not have a consistent correlation with the outcomes of qualitative studies. In this paper we analyze metrics of disentan…
Deconfounds neural network representation similarity metrics to improve consistency and accuracy.
problem Confounding by population structure in similarity metrics like RSA and CKA.
method Covariate adjustment regression to adjust for confounders.
result Improves detection of semantically similar neural networks and consistency in transfer learning.
This work extends holomorphic surface representations to isotropic space.
problem Representing minimal surfaces in simply isotropic space with degenerate metrics.
method Developed new forms of Weierstrass and Björling representations for isotropic minimal surfaces.
result Holomorphic representations of isotropic minimal surfaces are achieved.
New separation concepts for Anosov representations help bound Thurston asymmetric metric.
problem Understanding diverging families of Anosov representations.
method Introducing separation concepts and analyzing combinatorial invariants.
result Critical exponent asymptotic to a graph invariant.
Proposes a new metric to quantify the difference between neural network representations based on downstream task performance.
problem The lack of a consistent metric to measure the difference between neural network representations.
method Introduced the Transferred Discrepancy (TD) metric, which evaluates the difference between representations based on their performance on downstream tasks.
result TD provides fine-grained information for various downstream tasks and can evaluate the effectiveness of different training strategies.
A new method for representation and metric learning on manifolds boosts performance.
problem Improving representation and metric learning on complex, non-Euclidean data.
method Atlas-based manifold representation with a modified encoder and MMD loss.
result Substantial performance boost over baseline for low-dimensional encodings.
Geometric approach connects Burau representation to sphere metrics, identifying kernels.
problem Faithfulness of the Burau representation for the 4-strand case.
method Geometric and orbifold theory.
result Identifies the kernel of the Burau representation for some cases.
The dynamics of representations into PSL_d(R) are studied for surfaces of genus at least 3.
problem Dynamics of representations into PSL_d(R) for surfaces of genus at least 3.
method Showed quasi-convex subsets of infinite diameter for the Weil--Petersson metric have finite diameter for the path metric of the pressure metric through controlled bounded length of biinfinite paths of bending deformations.
result Biinfinite paths of bending deformations have controlled bounded length.
Riemannian metric learning improves data representation across various fields.
problem Traditional distance metrics fail to capture intrinsic data geometry.
method Leverages differential geometry to model data on Riemannian manifolds.
result Demonstrates remarkable success in diverse domains.
Deep networks are well-known to be fragile to adversarial attacks. We conduct an empirical analysis of deep representations under the state-of-the-art attack method called PGD, and find that the attack causes the internal representation to shift closer to the "false" class. Motivated by this observation, we propose to …
Optimal transport metric transfers deep network representations efficiently.
problem Efficiently transfer deep network representations for new tasks.
method Use optimal transport to quantify representation similarity and regularize student network.
result Optimal transport distance promotes similarity between teacher and student representations.
Geometric stability measures neural network robustness, distinguishing from similarity metrics.
problem Lack of robustness in neural network representations.
method Introduces geometric stability, quantified by Shesha metric measuring self-consistency.
result Stability and similarity are uncorrelated, revealing distinct properties of neural network robustness.
IsUMap improves data visualization of complex geometries.
problem Accurately representing complex, locally distorted metric spaces.
method Integrates UMAP and Isomap with Vietoris-Rips filtrations.
result Significant improvements in data representation quality.
Geodesic orbit metrics on real flag manifolds identified.
problem Classifying real flag manifolds with geodesic orbit metrics.
method Investigated invariant metrics on real flag manifolds, focusing on those where geodesics are orbits of one-parameter subgroups.
result Non-trivial geodesic orbit metrics exist on real flag manifolds, unlike in the complex case.
Reference metrics are used to define the differential structure on multicube representations of manifolds, i.e., they provide a simple and practical way to define what it means globally for tensor fields and their derivatives to be continuous. This paper introduces a general procedure for constructing reference metrics…
We study the existence of invariant Einstein metrics on real flag manifolds associated to simple and non-compact split real forms of complex classical Lie algebras whose isotropy representation decomposes into two or three irreducible sub-representations. In this situation, one can have equivalent sub-modules, leading …
Proposes a new metric learning method for image recognition.
problem Improving image recognition performance using learned distance representations.
method Introduces a Generalized Hybrid Metric Loss (GHM-Loss) to learn hybrid proximity features combining geometric and probabilistic spaces.
result Demonstrates superior performance compared to existing methods on public datasets.
A new embedding method extracts dataset-scale metric distribution into vectorial representation for graph data.
problem Classifying graph-structured data based on overall dataset-scale discrepancies.
method MetricDistribution2vec embedding strategy.
result Significant improvement in supervised prediction tasks on real-world graph datasets.
Study on Blaschke locus with covariance metric properties.
problem Riemannian structure of Blaschke locus.
method Analysis of Blaschke locus as a manifold, study of covariance metric properties.
result Geodesics in Blaschke locus have infinite length with respect to the covariance metric.
Learning rich representation from data is an important task for deep generative models such as variational auto-encoder (VAE). However, by extracting high-level abstractions in the bottom-up inference process, the goal of preserving all factors of variations for top-down generation is compromised. Motivated by the conc…
Proposes DWMD for better matching of hidden representations across domains.
problem Measuring data distribution discrepancy between semantically related domains for feature representation matching.
method DWMD, a moment-based probability distribution metric that explicitly orders and weights higher-order moments.
result DWMD is error-free and can strictly reflect distribution differences without feature distribution assumptions.
We propose a metric, Layer Saturation, defined as the proportion of the number of eigenvalues needed to explain 99% of the variance of the latent representations, for analyzing the learned representations of neural network layers. Saturation is based on spectral analysis and can be computed efficiently, making live ana…
Anti-de Sitter spacetimes embed cone-metrics as bent Cauchy surfaces.
problem Embedding cone-metrics in anti-de Sitter spacetimes.
method Proving embeddings using Fuchsian representations and GHMC spacetimes.
result Unique embeddings of cone-metrics in GHMC anti-de Sitter spacetimes.
New method removes unwanted information from representations efficiently.
problem Learning representations that are uninformative about a target variable.
method Adversarial training with a novel proxy metric for mutual information, leading to an analytically computable approximation.
result Our method effectively removes unwanted information with limited time budget.
New metric on geodesic currents connects different surface genera.
problem Understanding geodesic currents on surfaces of varying genera.
method Introducing a new asymmetric metric on the space of projective filling geodesic currents.
result Metric spaces of projective filling geodesic currents for surfaces of different genera are not isometric.
Every finite dimensional real representation of a compact real semisimple Lie algebra determines a metric 2-step nilpotent Lie algebra and a corresponding simply connected metric 2-step nilpotent Lie group N. We study the differential geometry of N using representation theory of the complexified complex semisimple Lie …
New proof shows holomorphic sectional curvature fully determines curvature tensor.
problem Determining the curvature tensor from holomorphic sectional curvature.
method Representation-theoretic means to calculate L2-norm of holomorphic sectional curvature. result Holomorphic sectional curvature fully determines the curvature tensor.
Study compares metrics from negative curvature and quasi-Fuchsian representations.
problem Comparing metrics on surface groups from negative curvature and quasi-Fuchsian representations.
method Examines Teichmüller space as the intersection of two metric families.
result Teichmüller space is the only common part of the two metric families.
This work characterizes how data augmentation shapes neural representations.
problem Understanding the impact of data augmentation on neural network representations.
method Embedding neural network hidden representations into a metric space invariant to transformations, analyzing shape-space trajectories.
result Increasing data augmentation strength leads to well-behaved trajectories in the embedded space, and different augmentation types steer representations in distinct directions.
Method learns representations invariant to task-irrelevant details in reinforcement learning tasks.
problem Learning representations that are invariant to task-irrelevant details in reinforcement learning.
method Uses bisimulation metrics to learn robust latent representations that encode only task-relevant information.
result Demonstrates SOTA performance in modified visual MuJoCo tasks and a first-person driving task.