arXiv subjects

Leonardo Concepción

Publications and source records attributed to Leonardo Concepción.

4 recordsLinked to original sources

arXiv2026

Lightweight CNN-Based Anomaly Detection for High Voltage Converter Modulators in the Spallation Neutron Source

Time series anomaly detection is central to monitoring industrial and scientific equipment, where anomalous sensor readings can precede failures. As such equipment usually carries several sensors, a detector must also decide how to model the dependencies between channels. In many physical systems, normal behaviour is defined by how groups of sensors evolve together, and an anomaly can appear as the coupling or decoupling of sensors while each signal remains normal. Anomalies in popular public benchmarks, however, are mostly visible in single channels, so these benchmarks cannot show whether modelling inter-channel dependencies improves detection. We study this on the High Voltage Converter Modulators (HVCMs) of the Spallation Neutron Source (SNS), whose fourteen sensors are physically coupled and whose faults are labelled by type. Keeping a convolutional neural network (CNN) fixed, we vary only whether its blocks combine channels before or after temporal filtering, and measure detection for each SNS subsystem and fault family. The effect of this choice depends on how visible each fault family is on a single channel. Families clearly visible on one channel are detected equally well by all variants, whereas weakly separable families, such as flux faults, gain consistently when channels are combined first. Modelling the dependencies between channels therefore helps only where individual channels do not already reveal the fault. The best detector reaches an average area under the precision--recall curve of 0.816, competitive with published HVCM detectors at roughly half the parameters of a standard CNN.

cs.LG↗

arXiv2026

Too much of a good thing -- when knowledge distillation promotes overfitting, and how to avoid it

The growing size of Convolutional Neural Networks has led to increasingly large and costly models. Knowledge Distillation (KD) addresses this by transferring knowledge from a large network (teacher) to a small one (student), also reducing the training data required. KD is traditionally applied only at the network's final output. However, its behaviour when applied at intermediate network layers has received little attention. This raises the question of whether intermediate block-wise KD, which provides supervision throughout the network, could offer an advantage under specific conditions, such as few instances per class, which is common in fine-grained datasets. This work proposes a student design based on simple, homogeneous blocks mirroring those of the teacher, distilling knowledge between corresponding blocks. Across eleven datasets, we show that on classic datasets, distilling only the last block is sufficient -- and often best--, whereas fine-grained, data-scarce settings benefit substantially from intermediate supervision, with even a single additional distillation point narrowing the gap considerably. We further study how this supervision should be guided, exploring configurations of varying granularity and informed by an explainability analysis based on attention maps, Centered Kernel Alignment, and Grad-CAM, alongside the impact of teacher and student fine-tuning strategies. This work shows that intermediate block-wise distillation, guided appropriately, is key to building compact data-efficient models without sacrificing accuracy.

cs.CV↗

arXiv2026

VACE: Learning Geometrically Structured Representations for Time Series Anomaly Detection

Anomaly detection in multivariate time series is a critical task across a wide range of real-world applications, where abnormal behaviour is rare, labels are unavailable, and the cost of a miss is high. The central challenge is learning a characterisation of normality precise enough to flag deviations. Representation self-supervised learning, typically through contrastive approaches, addresses this by embedding temporal patches into a latent space where normality occupies a well-defined region, with anomalies detected by geometric deviation. However, contrastive approaches shape this space indirectly through pair-sampling heuristics, providing no explicit control over the geometric structure that distance-based scoring requires. This means how tightly normal representations are grouped, and whether distances are directionally meaningful. We present VACE (Velocity-Aligned Channel Embeddings), a self-supervised anomaly detection method that represents normality as a compact, directionally coherent region in the embedding space. To this end, VACE trains a channel-aware encoder through a velocity-consistency objective, with no negatives and no synthetic anomalies, so that normal trajectories are locally smooth and aligned. At test time, a Mahalanobis positional score and a velocity-bank directional score are combined multiplicatively, flagging points that are simultaneously off-distribution and dynamically atypical. Despite its simplicity, VACE achieves state-of-the-art performance on TSB-AD-M under rigorous evaluation, significantly outperforming more complex methods trained on substantially larger budgets.

cs.LG↗

arXiv2022

Modeling Implicit Bias with Fuzzy Cognitive Maps

This paper presents a Fuzzy Cognitive Map model to quantify implicit bias in structured datasets where features can be numeric or discrete. In our proposal, problem features are mapped to neural concepts that are initially activated by experts when running what-if simulations, whereas weights connecting the neural concepts represent absolute correlation/association patterns between features. In addition, we introduce a new reasoning mechanism equipped with a normalization-like transfer function that prevents neurons from saturating. Another advantage of this new reasoning mechanism is that it can easily be controlled by regulating nonlinearity when updating neurons' activation values in each iteration. Finally, we study the convergence of our model and derive analytical conditions concerning the existence and unicity of fixed-point attractors.

cs.LG↗