14 ANITI papers presented at the 2022 NeurIPS conference

This year, 14 papers will be presented at the 2022 NeurIPS conference which will take place from novembre 28th to december 9th in New Orleans.

About the papers

• A benchmark for compositional visual reasoning. A. Zerroug, M. Vaishnav & J. Colin, S. Muslick & T. Serre. In: Neural Information Processing Systems (NeurIPS), 2022

• Aligning deep neural network strategies for object recognition with humans. T. Fel, I. F. Rodriguez, D. Linsley* & T. Serre. In: Neural Information Processing Systems (NeurIPS), 2022

• A Discourse on MetODS: Meta-optimized dynamical synapses for meta-reinforcement learning. M. Chalvidal, T. Serre & R. VanRullen.In: Neural Information Processing Systems (NeurIPS), 2022
  

• What I cannot predict, I do not understand: A human-centered evaluation framework for explainability methods. J. Colin, T. Fel, R. Cadene & T. Serre. In: Neural Information Processing Systems (NeurIPS), 2022

• Diversity vs. recognizability: Human-like generalization in one-shot generative models. V. Boutin, L. Singhal, X. Thomas & T. Serre.In: Neural Information Processing Systems (NeurIPS), 2022

• Local Identifiability of Deep ReLU Neural Networks: the Theory – Joachim Bona-Pellissier, François Malgouyres, Francois Bachoc – (NeurIPS), 2022

• High-dimensional Additive Gaussian Processes under Monotonicity Constraints – Andrés López-Lopera, Francois Bachoc, Olivier Roustant – (NeurIPS), 2022

• A general approximation lower bound in Lp norm, with applications to feed-forward neural networks – E. Achour, A. Foucault, S. Gerchinovitz, F. Malgouyres – (NeurIPS), 2022

• Automatic Differentiation of Nonsmooth Iterative Algorithms – J. Bolte, E. Pauwels, S. Vaiter – (NeurIPS), 2022

• Local Identifiability of Deep ReLU Neural Networks: the Theory – J. Bona-Pellissier, F. Malgouyres, F. Bachoc – (NeurIPS), 2022

• Pay attention to your loss: understanding misconceptions about 1-Lipschitz neural networks – Louis Béthune, Thibaut Boissin, Mathieu Serrurier, Franck Mamalet, Corentin Friedrich, Alberto González-Sanz

• Certificable Metric One Class Learning with Lipschitz Neural Networks – Louis Béthune, Mathieu Serrurier

• Making Sense of Dependence: Efficient Black-box Explanations Using Dependence Measure – Paul Novello Thomas Fel, David Vigouroux

• Look where you look! Saliency-guided Q-networks for visual RL tasks – David Bertoin, Adil Zouitine, Mehdi Zouitine, Emmanuel Rachelson

About NeurIPS

Founded in 1987, NeurIPS (Neural Information Processing Systems) is an international scientific conference in artificial intelligence and computational neuroscience.

The NeuIPS conference has always been focused on neural systems, whether we are talking about modeling biological neurons or artificial neurons. Over time, with the unprecedented development of digital technology, it has become the largest international meeting in statistical learning (machine learning) and in artificial intelligence.

This annual interdisciplinary and multi-disciplinary conference with distinguished speakers includes demonstrations, symposia, oral presentations, etc. It is accompanied by a professional exhibition focused on machine learning, and thematic workshops that provide a less formal setting for the exchange of ideas.