Výsledky vyhledávání - "Vilas, Martina"

Report

The Computational Complexity of Circuit Discovery for Inner Interpretability

Autor: Adolfi, Federico, Vilas, Martina G., Wareham, Todd

Many proposed applications of neural networks in machine learning, cognitive/brain science, and society hinge on the feasibility of inner interpretability via circuit discovery. This calls for empirical and theoretical explorations of viable algorith

Externí odkaz: http://arxiv.org/abs/2410.08025

Zobrazit plný text záznamu

Report

Limited but consistent gains in adversarial robustness by co-training object recognition models with human EEG

Autor: Guo, Manshan, Choksi, Bhavin, Sadiya, Sari, Gifford, Alessandro T., Vilas, Martina G., Cichy, Radoslaw M., Roig, Gemma

In contrast to human vision, artificial neural networks (ANNs) remain relatively susceptible to adversarial attacks. To address this vulnerability, efforts have been made to transfer inductive bias from human brains to ANNs, often by training the ANN

Externí odkaz: http://arxiv.org/abs/2409.03646

Zobrazit plný text záznamu

Report

Human-inspired Explanations for Vision Transformers and Convolutional Neural Networks

Autor: Panda, Mahadev Prasad, Tiezzi, Matteo, Vilas, Martina, Roig, Gemma, Eskofier, Bjoern M., Zanca, Dario

We introduce Foveation-based Explanations (FovEx), a novel human-inspired visual explainability (XAI) method for Deep Neural Networks. Our method achieves state-of-the-art performance on both transformer (on 4 out of 5 metrics) and convolutional mode

Externí odkaz: http://arxiv.org/abs/2408.02123

Zobrazit plný text záznamu

Report

Position: An Inner Interpretability Framework for AI Inspired by Lessons from Cognitive Neuroscience

Autor: Vilas, Martina G., Adolfi, Federico, Poeppel, David, Roig, Gemma

Inner Interpretability is a promising emerging field tasked with uncovering the inner mechanisms of AI systems, though how to develop these mechanistic theories is still much debated. Moreover, recent critiques raise issues that question its usefulne

Externí odkaz: http://arxiv.org/abs/2406.01352

Zobrazit plný text záznamu

Report

Analyzing Vision Transformers for Image Classification in Class Embedding Space

Autor: Vilas, Martina G., Schaumlöffel, Timothy, Roig, Gemma

Despite the growing use of transformer models in computer vision, a mechanistic understanding of these networks is still needed. This work introduces a method to reverse-engineer Vision Transformers trained to solve image classification tasks. Inspir

Externí odkaz: http://arxiv.org/abs/2310.18969

Zobrazit plný text záznamu

Report

Net2Brain: A Toolbox to compare artificial vision models with human brain responses

Autor: Bersch, Domenic, Dwivedi, Kshitij, Vilas, Martina, Cichy, Radoslaw M., Roig, Gemma

We introduce Net2Brain, a graphical and command-line user interface toolbox for comparing the representational spaces of artificial deep neural networks (DNNs) and human brain recordings. While different toolboxes facilitate only single functionaliti

Externí odkaz: http://arxiv.org/abs/2208.09677

Zobrazit plný text záznamu

Akademický článek

Tento výsledek nelze pro nepřihlášené uživatele zobrazit.
K zobrazení výsledku je třeba se přihlásit.

Vyhledávací nástroje:

Upřesnit hledání