Search

Scholarly Works (11 results)

Sort By:

Show:

Article
Peer Reviewed

Multiple Realizability and the Rise of Deep Learning

Proceedings of the Annual Meeting of the Cognitive Science Society, Volume 46 (2024)

The multiple realizability thesis holds that psychological states may be implemented in a diversity of physical systems. The deep learning revolution seems to be bringing this possibility to life, offering the most plausible examples of man-made realizations of sophisticated cognitive functions to date. This paper explores the implications of deep learning models for the multiple realizability thesis. Among other things, it challenges the widely held view that multiple realizability entails that the study of the mind can and must be pursued independently of the study of its implementation in the brain or in artificial analogues. Although its central contribution is philosophical, the paper has substantial methodological upshots for contemporary cognitive science, suggesting that deep neural networks may play a crucial role in formulating and evaluating hypotheses about cognition, even if they are interpreted as implementation-level models. In the age of deep learning, multiple realizability possesses a renewed significance.

Cover page: Multiple Realizability and the Rise of Deep Learning

Creative Commons 'BY' version 4.0 license

Thesis
Peer Reviewed

Compositionality and Cognitive Control in Neural Networks

Russin, Jacob
Advisor(s): O'Reilly, Randall C

UC Davis Electronic Theses and Dissertations (2023)

Compositionality, a natural property of symbolic systems, is thought to be a key principle underlying human intelligence: known concepts can be combined in novel ways according to systematic rules, allowing for the potentially infinite expressivity of human thought and language. Neural network models of cognition have long been criticized for failing to capture this important property. Despite their massive success in cognitive domains such as natural language processing, modern deep neural networks still struggle to generalize on compositional problems in the same ways that humans do, leading some to conclude that these networks must be augmented with symbolic or rule-like operations to fully account for key aspects of human cognition. Others have attempted to discover the inductive biases that would encourage compositionality to emerge in neural networks, without the need for explicit symbols or rules.The work presented in this dissertation takes the latter approach, exploring in particular the possibility that the mechanisms in the human brain responsible for cognitive control and top-down attentional modulation may constitute just such an inductive bias. Deep neural networks are used to study compositionality, cognitive control, and their relationship in both machine learning and cognitive neuroscience settings. Methods for measuring the extent to which compositional processing has emerged in neural networks are developed, and cognition-inspired attentional mechanisms are tested on compositional generalization problems. Additionally, neural networks are used to model phenomena observed with fMRI regarding control processes in the context of cognitive map formation.

Cover page: Compositionality and Cognitive Control in Neural Networks

Article
Peer Reviewed

Human Curriculum Effects Emerge with In-Context Learning in Neural Networks

Proceedings of the Annual Meeting of the Cognitive Science Society, Volume 46 (2024)

Human learning is sensitive to rule-like structure and the curriculum of examples used for training. In tasks governed by succinct rules, learning is more robust when related examples are blocked across trials, but in the absence of such rules, interleaving is more effective. To date, no neural model has simultaneously captured these seemingly contradictory effects. Here we show that this same tradeoff spontaneously emerges with ‚Äúin-context learning‚Äù (ICL) both in neural networks trained with metalearning and in large language models (LLMs). ICL is the ability to learn new tasks ‚Äúin context‚Äù ‚Äî without weight changes ‚Äî via an inner-loop algorithm implemented in activation dynamics. Experiments with pretrained LLMs and metalearning transformers show that ICL exhibits the blocking advantage demonstrated in humans on a task involving rule-like structure, and conversely, that concurrent in-weight learning reproduces the interleaving advantage observed in humans on tasks lacking such structure.

Cover page: Human Curriculum Effects Emerge with In-Context Learning in Neural Networks

Article
Peer Reviewed

Deep Predictive Learning in Neocortex and Pulvinar

UC Davis Previously Published Works (2021)

How do humans learn from raw sensory experience? Throughout life, but most obviously in infancy, we learn without explicit instruction. We propose a detailed biological mechanism for the widely embraced idea that learning is driven by the differences between predictions and actual outcomes (i.e., predictive error-driven learning). Specifically, numerous weak projections into the pulvinar nucleus of the thalamus generate top-down predictions, and sparse driver inputs from lower areas supply the actual outcome, originating in Layer 5 intrinsic bursting neurons. Thus, the outcome representation is only briefly activated, roughly every 100 msec (i.e., 10 Hz, alpha), resulting in a temporal difference error signal, which drives local synaptic changes throughout the neocortex. This results in a biologically plausible form of error backpropagation learning. We implemented these mechanisms in a large-scale model of the visual system and found that the simulated inferotemporal pathway learns to systematically categorize 3-D objects according to invariant shape properties, based solely on predictive learning from raw visual inputs. These categories match human judgments on the same stimuli and are consistent with neural representations in inferotemporal cortex in primates.

Cover page: Deep Predictive Learning in Neocortex and Pulvinar

Article
Peer Reviewed

Systematicity in a Recurrent Neural Network by Factorizing Syntax andSemantics

Proceedings of the Annual Meeting of the Cognitive Science Society, Volume 42 (2020)

Standard methods in deep learning fail to capture composi-tional or systematic structure in their training data, as shownby their inability to generalize outside of the training distribu-tion. However, human learners readily generalize in this way,e.g. by applying known grammatical rules to novel words. Theinductive biases that might underlie this powerful cognitive ca-pacity remain unclear. Inspired by work in cognitive sciencesuggesting a functional distinction between systems for syn-tactic and semantic processing, we implement a modificationto an existing deep learning architecture, imposing an analo-gous separation. The resulting architecture substantially out-performs standard recurrent networks on the SCAN dataset, acompositional generalization task, without any additional su-pervision. Our work suggests that separating syntactic fromsemantic learning may be a useful heuristic for capturing com-positional structure, and highlights the potential of using cog-nitive principles to inform inductive biases in deep learning.

Cover page: Systematicity in a Recurrent Neural Network by Factorizing Syntax andSemantics

Article
Peer Reviewed

How Sequential Interactive Processing Within Frontostriatal Loops Supports a Continuum of Habitual to Controlled Processing

UC Davis Previously Published Works (2020)

We address the distinction between habitual/automatic vs. goal-directed/controlled behavior, from the perspective of a computational model of the frontostriatal loops. The model exhibits a continuum of behavior between these poles, as a function of the interactive dynamics among different functionally-specialized brain areas, operating iteratively over multiple sequential steps, and having multiple nested loops of similar decision making circuits. This framework blurs the lines between these traditional distinctions in many ways. For example, although habitual actions have traditionally been considered purely automatic, the outer loop must first decide to allow such habitual actions to proceed. Furthermore, because the part of the brain that generates proposed action plans is common across habitual and controlled/goal-directed behavior, the key differences are instead in how many iterations of sequential decision-making are taken, and to what extent various forms of predictive (model-based) processes are engaged. At the core of every iterative step in our model, the basal ganglia provides a "model-free" dopamine-trained Go/NoGo evaluation of the entire distributed plan/goal/evaluation/prediction state. This evaluation serves as the fulcrum of serializing otherwise parallel neural processing. Goal-based inputs to the nominally model-free basal ganglia system are among several ways in which the popular model-based vs. model-free framework may not capture the most behaviorally and neurally relevant distinctions in this area.

Cover page: How Sequential Interactive Processing Within Frontostriatal Loops Supports a Continuum of Habitual to Controlled Processing

Article
Peer Reviewed

The Structure of Systematicity in the Brain

UC Davis Previously Published Works (2022)

A hallmark of human intelligence is the ability to adapt to new situations, by applying learned rules to new content (systematicity) and thereby enabling an open-ended number of inferences and actions (generativity). Here, we propose that the human brain accomplishes these feats through pathways in the parietal cortex that encode the abstract structure of space, events, and tasks, and pathways in the temporal cortex that encode information about specific people, places, and things (content). Recent neural network models show how the separation of structure and content might emerge through a combination of architectural biases and learning, and these networks show dramatic improvements in the ability to capture systematic, generative behavior. We close by considering how the hippocampal formation may form integrative memories that enable rapid learning of new structure and content representations.

Cover page: The Structure of Systematicity in the Brain

Article
Peer Reviewed

Complementary Structure-Learning Neural Networks for Relational Reasoning

Proceedings of the Annual Meeting of the Cognitive Science Society, Volume 43 (2021)

The neural mechanisms supporting flexible relational inferences, especially in novel situations, are a major focus of current research. In the complementary learning systems framework, pattern separation in the hippocampus allows rapid learning in novel environments, while slower learning in neocortex accumulates small weight changes to extract systematic structure from well-learned environments. In this work, we adapt this framework to a task from a recent fMRI experiment where novel transitive inferences must be made according to implicit relational structure. We show that computational models capturing the basic cognitive properties of these two systems can explain relational transitive inferences in both familiar and novel environments, and reproduce key phenomena observed in the fMRI experiment.

Cover page: Complementary Structure-Learning Neural Networks for Relational Reasoning

Article
Peer Reviewed

Compositional processing emerges in neural networks solving math problems

Proceedings of the Annual Meeting of the Cognitive Science Society, Volume 43 (2021)

A longstanding question in cognitive science concerns the learning mechanisms underlying compositionality in human cognition. Humans can infer the structured relationships (e.g., grammatical rules) implicit in their sensory observations (e.g., auditory speech), and use this knowledge to guide the composition of simpler meanings into complex wholes. Recent progress in artificial neural networks has shown that when large models are trained on enough linguistic data, grammatical structure emerges in their representations. We extend this work to the domain of mathematical reasoning, where it is possible to formulate precise hypotheses about how meanings (e.g., the quantities corresponding to numerals) should be composed according to structured rules (e.g., order of operations). Our work shows that neural networks are not only able to infer something about the structured relationships implicit in their training data, but can also deploy this knowledge to guide the composition of individual meanings into composite wholes.

Cover page: Compositional processing emerges in neural networks solving math problems

Article
Peer Reviewed

A Neural Network Model of Continual Learning with Cognitive Control

Proceedings of the Annual Meeting of the Cognitive Science Society, Volume 44 (2022)

Neural networks struggle in continual learning settings from catastrophic forgetting: when trials are blocked, new learning can overwrite the learning from previous blocks. Humans learn effectively in these settings, in some cases even showing an advantage of blocking, suggesting the brain contains mechanisms to overcome this problem. Here, we build on previous work and show that neural networks equipped with a mechanism for cognitive control do not exhibit catastrophic forgetting when trials are blocked. We further show an advantage of blocking over interleaving when there is a bias for active maintenance in the control signal, implying a tradeoff between maintenance and the strength of control. Analyses of map-like representations learned by the networks provided additional insights into these mechanisms. Our work highlights the potential of cognitive control to aid continual learning in neural networks, and offers an explanation for the advantage of blocking that has been observed in humans.

Cover page: A Neural Network Model of Continual Learning with Cognitive Control