Papers Explored

2025

1. Understanding Model Calibration - A gentle introduction and visual exploration of calibration and the expected calibration error (ECE)

   Maja Pavlovic

   In ICLR Blogposts 2025, 2025

   https://iclr-blogposts.github.io/2025/blog/calibration/

   Abs HTML

   To be considered reliable, a model must be calibrated so that its confidence in each decision closely reflects its true outcome. In this blogpost we’ll take a look at the most commonly used definition for calibration and then dive into a frequently used evaluation measure for model calibration. We’ll then cover some of the drawbacks of this measure and how these surfaced the need for additional notions of calibration, which require their own new evaluation measures. This post is not intended to be an in-depth dissection of all works on calibration, nor does it focus on how to calibrate models. Instead, it is meant to provide a gentle introduction to the different notions and their evaluation measures as well as to re-highlight some issues with a measure that is still widely used to evaluate calibration.

2. Expressive Power of Temporal Message Passing

   Przemysław Andrzej Wałęga and Michael Rawson

   Proceedings of the AAAI Conference on Artificial Intelligence, Apr 2025

   DOI
3. Deep Learning is Not So Mysterious or Different

   Andrew Gordon Wilson

   Apr 2025

   HTML

2024

1. NEURAL NETWORK COMPRESSION: THE FUNCTIONAL PERSPECTIVE

   Israel Mason-Williams

   In 5th Workshop on practical ML for limited/low resource settings, Apr 2024

   HTML
2. Knowledge Distillation: The Functional Perspective

   Israel Mason-Williams, Gabryel Mason-Williams, and Mark Sandler

   In NeurIPS 2024 Workshop on Scientific Methods for Understanding Deep Learning, Apr 2024

   HTML
3. Expressive Power of Temporal Message Passing

   Przemysław Andrzej Wałęga and Michael Rawson

   Apr 2024

   HTML
4. Approaching Deep Learning through the Spectral Dynamics of Weights

   David Yunis, Kumar Kshitij Patel, Samuel Wheeler, and 5 more authors

   Apr 2024

   HTML
5. Dynamical mechanisms of how an RNN keeps a beat, uncovered with a low-dimensional reduced model

   Klavdia Zemlianova, Amitabha Bose, and John Rinzel

   Scientific Reports, Apr 2024

   HTML
6. Flexible multitask computation in recurrent networks utilizes shared dynamical motifs

   Laura N Driscoll, Krishna Shenoy, and David Sussillo

   Nature Neuroscience, Apr 2024

   HTML
7. What Makes a Good Prune? Maximal Unstructured Pruning for Maximal Cosine Similarity

   Gabryel Mason-Williams and Fredrik Dahlqvist

   In The Twelfth International Conference on Learning Representations, Apr 2024

   HTML

2023

1. Hebbian learning with elasticity explains how the spontaneous motor tempo affects music performance synchronization

   Iran R Roman, Adrian S Roman, Ji Chul Kim, and 1 more author

   PLOS Computational Biology, Apr 2023

   HTML

2021

1. Predicting trends in the quality of state-of-the-art neural networks without access to training or testing data

   Charles H Martin, Tongsu Peng, and Michael W Mahoney

   Nature Communications, Apr 2021

   HTML

2020

1. Understanding generalization through visualizations

   W Ronny Huang, Zeyad Emam, Micah Goldblum, and 4 more authors

   Apr 2020

   HTML

2019

1. Deep ensembles: A loss landscape perspective

   Stanislav Fort, Huiyi Hu, and Balaji Lakshminarayanan

   arXiv preprint arXiv:1912.02757, Apr 2019

   HTML

2018

1. Measuring the Intrinsic Dimension of Objective Landscapes

   Chunyuan Li, Heerad Farkhoor, Rosanne Liu, and 1 more author

   In International Conference on Learning Representations, Apr 2018

   HTML

2013

1. Context-dependent computation by recurrent dynamics in prefrontal cortex

   Valerio Mante, David Sussillo, Krishna V Shenoy, and 1 more author

   nature, Apr 2013

   HTML

1989

1. Finite state automata and simple recurrent networks

   Axel Cleeremans, David Servan-Schreiber, and James L McClelland

   Neural computation, Apr 1989

   HTML