[HTML][HTML] Understanding adversarial examples requires a theory of artefacts for deep learning
C Buckner - Nature Machine Intelligence, 2020 - nature.com
Deep neural networks are currently the most widespread and successful technology in
artificial intelligence. However, these systems exhibit bewildering new vulnerabilities: most …
artificial intelligence. However, these systems exhibit bewildering new vulnerabilities: most …
[HTML][HTML] Solving the black box problem: A normative framework for explainable artificial intelligence
C Zednik - Philosophy & technology, 2021 - Springer
Many of the computing systems programmed using Machine Learning are opaque: it is
difficult to know why they do what they do or how they work. Explainable Artificial …
difficult to know why they do what they do or how they work. Explainable Artificial …
Performance vs. competence in human–machine comparisons
C Firestone - Proceedings of the National Academy of …, 2020 - National Acad Sciences
Does the human mind resemble the machines that can behave like it? Biologically inspired
machine-learning systems approach “human-level” accuracy in an astounding variety of …
machine-learning systems approach “human-level” accuracy in an astounding variety of …
Understanding from machine learning models
E Sullivan - The British Journal for the Philosophy of Science, 2022 - journals.uchicago.edu
Simple idealized models seem to provide more understanding than opaque, complex, and
hyper-realistic models. However, an increasing number of scientists are going in the …
hyper-realistic models. However, an increasing number of scientists are going in the …
The rhetoric and reality of anthropomorphism in artificial intelligence
D Watson - Minds and Machines, 2019 - Springer
Artificial intelligence (AI) has historically been conceptualized in anthropomorphic terms.
Some algorithms deploy biomimetic designs in a deliberate attempt to effect a sort of digital …
Some algorithms deploy biomimetic designs in a deliberate attempt to effect a sort of digital …
Two dimensions of opacity and the deep learning predicament
FJ Boge - Minds and Machines, 2022 - Springer
Deep neural networks (DNNs) have become increasingly successful in applications from
biology to cosmology to social science. Trained DNNs, moreover, correspond to models that …
biology to cosmology to social science. Trained DNNs, moreover, correspond to models that …
Situated neural representations: Solving the problems of content
G Piccinini - Frontiers in Neurorobotics, 2022 - frontiersin.org
Situated approaches to cognition maintain that cognition is embodied, embedded, enactive,
and affective (and extended, but that is not relevant here). Situated approaches are often …
and affective (and extended, but that is not relevant here). Situated approaches are often …
Deep learning: A philosophical introduction
C Buckner - Philosophy compass, 2019 - Wiley Online Library
Deep learning is currently the most prominent and widely successful method in artificial
intelligence. Despite having played an active role in earlier artificial intelligence and neural …
intelligence. Despite having played an active role in earlier artificial intelligence and neural …
Beyond generalization: a theory of robustness in machine learning
T Freiesleben, T Grote - Synthese, 2023 - Springer
The term robustness is ubiquitous in modern Machine Learning (ML). However, its meaning
varies depending on context and community. Researchers either focus on narrow technical …
varies depending on context and community. Researchers either focus on narrow technical …
[HTML][HTML] Methods for identifying emergent concepts in deep neural networks
T Räz - Patterns, 2023 - cell.com
The present perspective discusses methods to detect concepts in internal representations
(hidden layers) of deep neural networks (DNNs), such as network dissection, feature …
(hidden layers) of deep neural networks (DNNs), such as network dissection, feature …