arXiv research

Smooth calibration improves forecast reliability even with leaked information.

In this article we revisit the definition of Precision-Recall (PR) curves for generative models proposed by Sajjadi et al. (arXiv:1806.00035). Rather than providing a scalar for generative quality, PR curves distinguish mode-collapse (poor recall) and bad quality (poor precision). We first generalize their formulation …

Transformers learn to recall with non-orthogonal embeddings in realistic settings.

New algorithm reduces online learning regret for bounded recall games.

This paper introduces a new method to train normalizing flows using precision-recall divergences.

The study examines how class imbalance affects precision-recall curves.

In this paper, we show that the class of all properly 3-realizable groups is closed under amalgamated free products (and HNN-extensions) over finite groups. We recall that $G$ is said to be properly 3-realizable if there exists a compact 2-polyhedron $K$ with $π_1(K) \cong G$ and whose universal cover $\tilde{K}$ has t…

Cochran, Orr and Teichner introduced $L^2$--eta--invariants to detect highly non--trivial examples of non slice knots. Using a recent theorem by Lück and Schick we show that their metabelian $L^2$--eta--invariants can be viewed as the limit of finite dimensional unitary representations. We recall a ribbon obstruction t…

Proposes a method to evaluate classifiers with missing labels using multiple imputation.

The study formalizes temporal precision and recall for anomaly detection in sequences.

Three RFF-based methods for nonlinear causal discovery in mixed data.

For information retrieval and binary classification, we show that precision at the top (or precision at k) and recall at the top (or recall at k) are maximised by thresholding the posterior probability of the positive class. This finding is a consequence of a result on constrained minimisation of the cost-sensitive exp…

Unified and extended precision-recall metrics for generative models.

Transformers can store facts efficiently using associative memories.

How different is the universal cover of a given finite 2-complex from a 3-manifold (from the proper homotopy viewpoint)? Regarding this question, we recall that a finitely presented group $G$ is said to be properly 3-realizable if there exists a compact 2-polyhedron $K$ with $π_1(K) \cong G$ whose universal cover $\til…

Recalls intrinsically harmonic forms and open problems.

LLMs can memorize economic data and recall exact values before their training cutoff.

Paper tackles imbalanced binary classification by optimizing precision and recall directly.

RAGuard improves safety in LLMs for offshore wind maintenance.

Paper proposes a human-algorithm approach to reduce medical device recall risk and workload.

Framework for precise recall control in spatial conflation tasks.

A new method for generating replay samples on the fly, optimizing for not forgetting.

Proposes a new tree-based algorithm for class-imbalanced data.

Recall that a pseudofree group action on a space is one whose singular set consists only of isolated points. In this paper, we classify all of the finite groups which admit pseudofree actions on S^2 x S^2. The groups turn out to be exactly the expected ones: those which admit orthogonal pseudofree actions on S^2 x S^2 …

The paper critiques and expands on common evaluation metrics in machine learning.

In this brief sequel to a previous article, we recall the notion of a cut cellular surface (CCS), being a surface with boundary, which is cut in a specified way to be represented in the plane, and is composed of 0-, 1- and 2-cells. We obtain invariants of CCS's under Pachner-like moves on the cellular structure, by cou…

Interpreting neural network decisions and the information learned in intermediate layers is still a challenge due to the opaque internal state and shared non-linear interactions. Although (Kim et al, 2017) proposed to interpret intermediate layers by quantifying its ability to distinguish a user-defined concept (from r…

New findings show tool-augmented models can recall unlimited facts, outperforming purely memorized models.

The optimal ranking score between precision and recall is rarely F1 and can be found using specific methods.

We study the effectiveness of several techniques to personalize end-to-end speech models and improve the recognition of proper names relevant to the user. These techniques differ in the amounts of user effort required to provide supervision, and are evaluated on how they impact speech recognition performance. We propos…

The paper defines a new Lie groupoid concept for infinite dimensions.

Recalls and refines the concept of algebraically rectifiable curves.

DIAL learns embeddings to maximize recall and accuracy for entity resolution.

Transformers recall from long distributions with statistical guarantees.

We give a survey of the theory of surface braid groups and the lower algebraic K-theory of their group rings. We recall several definitions and describe various properties of surface braid groups, such as the existence of torsion, orderability, linearity, and their relation both with mapping class groups and with the h…

Short sales are regarded as negative purchases in textbook asset pricing theory. In reality, however, the symmetry between purchases and short sales is broken by a variety of costs and risks peculiar to the latter. We formulate an optimal stopping model in which the decision to cover a short position is affected by two…

The paper explores heat flow and constants on graphs, proving properties and proposing new concepts.

Despite the tremendous progress in the estimation of generative models, the development of tools for diagnosing their failures and assessing their performance has advanced at a much slower pace. Recent developments have investigated metrics that quantify which parts of the true distribution is modeled well, and, on the…

Classifies minimal surfaces of finite genus in 3D space.

Enhanced anomaly detection using PRC-RF with autoencoders.

LLMs can be tricked into recalling facts based on context clues.

We describe the Cartan and Weil models of twisted equivariant cohomology together with the Cartan homomorphism among the two, and we extend the Chern-Weil homomorphism to the twisted equivariant cohomology. We clarify that in order to have a cohomology theory, the coefficients of the twisted equivariant cohomology must…

Evaluation often aims to reduce the correctness or error characteristics of a system down to a single number, but that always involves trade-offs. Another way of dealing with this is to quote two numbers, such as Recall and Precision, or Sensitivity and Specificity. But it can also be useful to see more than this, and …

Unified framework for sequence models using test-time regression.

New method detects novel node categories in graphs with distribution shifts.

KATA improves associative recall by optimizing feature maps derived from nonnegative attention weights.

Constructing $3$-Lie algebroids via connections

This paper aims to provide a description of totally isotropic Willmore two-spheres and their adjoint transforms. We first recall the isotropic harmonic maps which are introduced by Hélein, Xia-Shen and Ma for the study of Willmore surfaces. Then we derive a description of the normalized potential (some Lie algebra valu…