L'actualité de l'IA, puisée à la source.

Le plus récent

arXiv — Machine Learning (cs.LG)Presse & veilleEN

Process then Retrieve: Investigating the Origins of the Modality Gap in Vision-Language Models (s'ouvre sur arXiv — Machine Learning (cs.LG), nouvel onglet)

arXiv:2611.09897v1 Announce Type: new Abstract: Understanding the origins of the modality gap in vision language models remains a significant challenge. This work investigates whether differences in information density between modality specific tokens at the embedding layer contr…

10 nov. 2026Lire l'article

1. 

   arXiv — Machine Learning (cs.LG)Presse & veilleEN

   MAGE: Towards Generalizable Multi-timescale EEG Representations (s'ouvre sur arXiv — Machine Learning (cs.LG), nouvel onglet)

   arXiv:2611.09928v1 Announce Type: new Abstract: Electroencephalography (EEG) data plays an important role in understanding human brain activity, with various major applications. While current approaches, which largely build on EEG foundation models, have demonstrated strong perfo…

   10 nov. 2026Lire l'article
2. 

   arXiv — Machine Learning (cs.LG)Presse & veilleEN

   Causal Discovery Bench: Can LLMs Autonomously Conduct Effective Experiments? (s'ouvre sur arXiv — Machine Learning (cs.LG), nouvel onglet)

   arXiv:2611.09934v1 Announce Type: new Abstract: As large language model (LLM) agents get actively adopted for autonomous research, they must not only reason over existing evidence, but also design informative experiments, allocate resources, analyze sequential observations, and d…

   10 nov. 2026Lire l'article
3. 

   arXiv — Machine Learning (cs.LG)Presse & veilleEN

   Fresh Enough to Decide: Age of Intelligence for Cost-Aware Model Synchronization in Dynamic IoT Systems (s'ouvre sur arXiv — Machine Learning (cs.LG), nouvel onglet)

   arXiv:2611.09935v1 Announce Type: new Abstract: Resource-constrained IoT devices cache learned models for real-time decisions, but frequent updates incur substantial bandwidth, energy, and service overhead. Because decision-relevant model differences need not grow with elapsed ti…

   10 nov. 2026Lire l'article
4. 

   arXiv — Machine Learning (cs.LG)Presse & veilleEN

   Topological Deep Learning Challenge 2026: Bridging the Gap (s'ouvre sur arXiv — Machine Learning (cs.LG), nouvel onglet)

   arXiv:2611.10046v1 Announce Type: new Abstract: This paper describes the 2026 edition of the Topological Deep Learning Challenge: Bridging the Gap, hosted at the second Topology, Algebra, and Geometry in Data Science (TAG-DS) Conference. Where previous editions grew the modelling…

   10 nov. 2026Lire l'article
5. 

   arXiv — Machine Learning (cs.LG)Presse & veilleEN

   Selected-Risk Calibration for Feasibility at Returned Decisions (s'ouvre sur arXiv — Machine Learning (cs.LG), nouvel onglet)

   arXiv:2611.10050v1 Announce Type: new Abstract: When an optimization problem uses a surrogate constraint, a decision that is feasible for the surrogate need not be feasible for the true constraint. We propose Selected-Risk Calibration (SRC), a framework that calibrates a learned …

   10 nov. 2026Lire l'article
6. 

   arXiv — Machine Learning (cs.LG)Presse & veilleEN

   All roads lead to Markov kernel networks: A representation theorem for discretization-stable learning (s'ouvre sur arXiv — Machine Learning (cs.LG), nouvel onglet)

   arXiv:2611.10051v1 Announce Type: new Abstract: Operator learning seeks to approximate function-to-function maps in a way that remains stable when the input discretization changes, a property sometimes called "discretization invariance". We consider a quantitative refinement of t…

   10 nov. 2026Lire l'article
7. 

   arXiv — Machine Learning (cs.LG)Presse & veilleEN

   fNIRSAtlas: A Large-Scale Benchmark for Functional Near-Infrared Spectroscopy Classification (s'ouvre sur arXiv — Machine Learning (cs.LG), nouvel onglet)

   arXiv:2611.10082v1 Announce Type: new Abstract: Functional near-infrared spectroscopy (fNIRS) is a promising modality for brain-computer interfaces and cognitive state decoding, but progress in machine learning for fNIRS has been limited by small, isolated datasets and inconsiste…

   10 nov. 2026Lire l'article

8 articles