arXiv subjects

Jeffrey Bowers

Publications and source records attributed to Jeffrey Bowers.

5 recordsLinked to original sources

arXiv2023

Convolutional Neural Networks Trained to Identify Words Provide a Surprisingly Good Account of Visual Form Priming Effects

A wide variety of orthographic coding schemes and models of visual word identification have been developed to account for masked priming data that provide a measure of orthographic similarity between letter strings. These models tend to include hand-coded orthographic representations with single unit coding for specific forms of knowledge (e.g., units coding for a letter in a given position). Here we assess how well a range of these coding schemes and models account for the pattern of form priming effects taken from the Form Priming Project and compare these findings to results observed with 11 standard deep neural network models (DNNs) developed in computer science. We find that deep convolutional networks (CNNs) perform as well or better than the coding schemes and word recognition models, whereas transformer networks did less well. The success of CNNs is remarkable as their architectures were not developed to support word recognition (they were designed to perform well on object recognition), they classify pixel images of words (rather than artificial encodings of letter strings), and their training was highly simplified (not respecting many key aspects of human experience). In addition to these form priming effects, we find that the DNNs can account for visual similarity effects on priming that are beyond all current psychological models of priming. The findings add to the recent work of (Hannagan et al., 2021) and suggest that CNNs should be given more attention in psychology as models of human visual word recognition.

cs.CV↗

arXiv2020

A case for robust translation tolerance in humans and CNNs. A commentary on Han et al

Han et al. (2020) reported a behavioral experiment that assessed the extent to which the human visual system can identify novel images at unseen retinal locations (what the authors call "intrinsic translation invariance") and developed a novel convolutional neural network model (an Eccentricity Dependent Network or ENN) to capture key aspects of the behavioral results. Here we show that their analysis of behavioral data used inappropriate baseline conditions, leading them to underestimate intrinsic translation invariance. When the data are correctly interpreted they show near complete translation tolerance extending to 14° in some conditions, consistent with earlier work (Bowers et al., 2016) and more recent work Blything et al. (in press). We describe a simpler model that provides a better account of translation invariance.

q-bio.NC↗

arXiv2020

Learning Translation Invariance in CNNs

When seeing a new object, humans can immediately recognize it across different retinal locations: we say that the internal object representation is invariant to translation. It is commonly believed that Convolutional Neural Networks (CNNs) are architecturally invariant to translation thanks to the convolution and/or pooling operations they are endowed with. In fact, several works have found that these networks systematically fail to recognise new objects on untrained locations. In this work we show how, even though CNNs are not 'architecturally invariant' to translation, they can indeed 'learn' to be invariant to translation. We verified that this can be achieved by pretraining on ImageNet, and we found that it is also possible with much simpler datasets in which the items are fully translated across the input canvas. We investigated how this pretraining affected the internal network representations, finding that the invariance was almost always acquired, even though it was some times disrupted by further training due to catastrophic forgetting/interference. These experiments show how pretraining a network on an environment with the right 'latent' characteristics (a more naturalistic environment) can result in the network learning deep perceptual rules which would dramatically improve subsequent generalization.

cs.CV↗

arXiv2019

Training neural networks to encode symbols enables combinatorial generalization

Combinatorial generalization - the ability to understand and produce novel combinations of already familiar elements - is considered to be a core capacity of the human mind and a major challenge to neural network models. A significant body of research suggests that conventional neural networks can't solve this problem unless they are endowed with mechanisms specifically engineered for the purpose of representing symbols. In this paper we introduce a novel way of representing symbolic structures in connectionist terms - the vectors approach to representing symbols (VARS), which allows training standard neural architectures to encode symbolic knowledge explicitly at their output layers. In two simulations, we show that neural networks not only can learn to produce VARS representations, but in doing so they achieve combinatorial generalization in their symbolic and non-symbolic output. This adds to other recent work that has shown improved combinatorial generalization under specific training conditions, and raises the question of whether specific mechanisms or training routines are needed to support symbolic processing.

cs.AI↗

arXiv2001

Crystalline Color Superconductivity

In any context in which color superconductivity arises in nature, it is likely to involve pairing between species of quarks with differing chemical potentials. For suitable values of the differences between chemical potentials, Cooper pairs with nonzero total momentum are favored, as was first realized by Larkin, Ovchinnikov, Fulde and Ferrell (LOFF). Condensates of this sort spontaneously break translational and rotational invariance, leading to gaps which vary periodically in a crystalline pattern. Unlike the original LOFF state, these crystalline quark matter condensates include both spin zero and spin one Cooper pairs. We explore the range of parameters for which crystalline color superconductivity arises in the QCD phase diagram. If in some shell within the quark matter core of a neutron star (or within a strange quark star) the quark number densities are such that crystalline color superconductivity arises, rotational vortices may be pinned in this shell, making it a locus for glitch phenomena.

hep-ph↗