16 citations · 58 across the 13 of their papers we have counts for
16 papers
The Blind Spot Paradox: When Adaptive Classifiers Defeat Drift Detectors
Raphaël Minato, Fabrice Popineau, Arpad Rimmel +1
Monitoring concept drift from an adaptive classifier's error stream creates an operational conflict with the model's own update loop. When internal adaptation outpaces evidence acc…
Mitigating Convergence Collapse in Fixed-Target Anomaly Detectors via Kernel-Anchored Locality Regularization
José Lucas De Melo Costa, Fabrice Popineau, Arpad Rimmel +1
A family of tabular anomaly detectors trains a neural map toward a fixed target under squared-error loss and scores anomalies by the test-time residual; contraction matching, one-s…
Drive vs. Decay: On the Training Dynamics of Joint-Embedding Predictive Architectures
José Lucas De Melo Costa, Seong Woo Ahn, Fabrice Popineau +2
Joint-Embedding Predictive Architectures (JEPAs) are prone to representation collapse, typically mitigated through empirical heuristics. We develop an early-training stability theo…
Knowledge-Informed Local Causal Discovery of Optimal Adjustment Sets
Seong Woo Ahn, Alessandro Leite, José Lucas De Melo Costa +3
Local causal discovery is a scalable alternative to global structure learning. However, it can struggle to identify valid adjustment sets in data-scarce settings because of finite-…
High Performance, Low Reliability: Uncertainty Benchmarking for Tabular Foundation Models
José Lucas De Melo Costa, Fabrice Popineau, Arpad Rimmel +1
Recent Tabular Foundation Models (TFMs) have demonstrated state-of-the-art predictive performance, often surpassing Gradient-Boosted Decision Trees (GBDTs). However, the trustworth…
T-JEPA: Augmentation-Free Self-Supervised Learning for Tabular Data
Hugo Thimonier, José Lucas De Melo Costa, Fabrice Popineau +2
Self-supervision is often used for pre-training to foster performance on a downstream task by constructing meaningful representations of samples. Self-supervised learning (SSL) gen…