Canada

Hippocampus (HPC) is a key structure in category learning, forming organized concepts from individual episodes to adaptively represent shared and distinct elements. Most of the evidence for the role of HPC in concept formation, however, relies on neural coding measured at the end of learning. Thus, it remains unclear how neural mechanisms early in learning support the formation of flexible knowledge. We explored this by combining human functional MRI with a computational model of learning, SUSTAIN. Participants learned to categorize complex visual objects with multiple features. A multidimensional rule linked features to categories with exception items in each category. We observed distinct HPC engagement along its primary axis for exception learning that also predicted learning outcomes. The model predicted that better learners use exceptions to update their knowledge, which was reflected in HPC activation. These findings suggest that HPC discriminates initial informative experiences during learning to support flexible knowledge formation.

CogSci 2022

Hippocampus Engagement Early in Learning Supports Concept Formation

poster

Join us for CogSci2022! Our 44th Annual Conference will be centered around the theme of diversity, with an accompanying range of keynotes and invited speakers who themselves highlight the diversity of our field. Our three symposia are Cognition across Diverse Populations, Cognition within Diverse Societies, and Global Challenges for Cognitive Science. The conference is being offered in a fully hybrid environment for onsite and online attendees.
 
 [**The Program at a Glance can be found here**](https://cognitivesciencesociety.org/wp-content/uploads/2022/07/Cogsci-2022-PAAG-Detailed-July-12final-1.pdf)
 [**The Detailed Program can be found here**](https://cognitivesciencesociety.org/wp-content/uploads/2022/08/DeArmond-CogSci2022-program-v113-final.pdf)

This conference requires a registration to attend. Please use the link below to secure your ticket. The website will be available to all registered attendees on July 25th. 

[Register to Attend](https://cognitivesciencesociety.org/registration/)

Registration Is Required

The 44th Annual Meeting of the Cognitive Science Society highlighted research on the theme Cognitive Diversity in addition to the full breadth and diversity of research topics offered by the society’s membership. 





Although the relationship between the sound of each word and its meaning is generally arbitrary, onomatopoeias are said to have the unarbitrary link, which called sound-symbolism, between them. In this study, we investigated whether sound symbolic words are widely common in a class of onomatopoeias in some natural language. We conducted an experiment, which asked Japanese and non-Japanese speakers to match each given Japanese-onomatopoeia-like sound with a shape to which the sound referred. The result of analysis showed a similar structure for both speaker groups, in which round shapes were associated to a particular set of sounds and pointed shapes associated to the other set of sounds. Moreover, the round/ pointed shapes are correlated to pseudo onomatopoeias with sonorants/ fricative phonetic features. This finding supports the sound symbolic hypothesis asserting that the major component of Japanese onomatopoeias forms a bouba-kiki like sound-shape correspondence even for non-Japanese speakers.

Identifying a Phonetic Factors of Onomatopoeias Correlated to Sound Symbolic Commons between Japanese and Non-Japanese Speakers

Data were collected from a brain-computer interface speller that utilized the P3b as a control signal. Stimuli consisted of letters and their "segments". Importantly, different letters were made up of different numbers of segments from a 10 segment library. Subjects were instructed to mentally note whenever segments from their letter (targets) were flashed. We found P3b amplitudes of target segments decreased as the number of segments in a letter (target letter complexity) increased. In contrast, the P3b attenuation was not affected by the total number of letters a segment belonged to (segment frequency). These results may reflect higher task difficulty caused by increased working memory load with increased target letter complexity. Alternatively, it's possible that despite the target rate being fixed at 30% within each block, subjects erroneously believed the target rate increased with target letter complexity. Further work to disentangle these possibilities may enrich our understanding of the P3b.

Set Size Effects on the P3b in a BCI Speller

Referring expression understanding and generation are critical for robots to communicate about the world around them. Recently there have been significant advances on the problem of referring expression understanding, also known as reference resolution, with researchers presenting approaches to both incremental reference resolution (i.e., processing referring expressions word by word in real-time as they are spoken) and open-world reference resolution (i.e., resolving references both to known and previously unknown entities). In this work, we combine insights from these approaches to present IPOWER: the first algorithm for performing reference resolution incrementally in open-world environments.

IPOWER: Incremental, Probabilistic, Open-World Reference Resolution

Systematic generalization is the ability to combine known parts into novel meaning; an important aspect of efficient human learning, but a weakness of neural network learning. In this work, we investigate how two well-known modeling principles --- modularity and data augmentation --- affect systematic generalization of neural networks in grounded language learning. We analyze how large the vocabulary needs to be to achieve systematic generalization and how similar the augmented data needs to be to the problem at hand. Our findings show that even in the controlled setting of a synthetic benchmark, achieving systematic generalization remains very difficult. After training on an augmented dataset with almost forty times more adverbs than the original problem, a non-modular baseline is not able to systematically generalize to a novel combination of a known verb and adverb. When separating the task into cognitive processes like perception and navigation, a modular neural network is able to utilize the augmented data and generalize more systematically, achieving 70% and 40% exact match increase over state-of-the-art on two gSCAN tests that have not previously been improved. We hope that this work gives insight into the drivers of systematic generalization, and what we still need to improve for neural networks to learn more like humans do.

Improving Systematic Generalization Through Modularity and Augmentation

Biological brains learn much more quickly than standard deep neural network reinforcement learning algorithms. One reason for this is that the deep neural networks need to learn a representation that is appropriate for the task at hand, whilst biological systems already possess an appropriate representation. Here, we bypass this problem by imposing on the neural network a representation based on what is observed in biology, such as grid cells. This study explores the impact of using a biologically-inspired grid-cell representation vs. a one-hot representation, on the speed at which a Temporal Difference-based Actor-Critic network learns to solve a simple 2D grid-world reinforcement learning task. The results suggest that the use of grid cells does promote faster learning. Furthermore, the grid cells implemented here have the potential for accurately representing unbounded continuous space. Thus, their promising performance on this discrete task acts as a first step in exploring their utility for reinforcement learning in continuous space.

Biologically-Based Neural Representations Enable Fast Online Shallow Reinforcement Learning

How do humans find their way in new environments, so quickly and efficiently? Humans reuse old knowledge to build new concepts in many non-spatial domains, such as language and drawing. Could people learn maps by a similar process, that extracts common structure to speed up learning, and generalize across maps? Understanding the computational cognitive mechanisms that support this efficiency can advance the study of the human mind and enable more efficient exploration algorithms. We hypothesize that human map learning relies on inferences over the structure of unobserved spaces, based on spatial priors informed by previous experience. We model this by combining Program Induction with a Hierarchical Bayesian framework that explicitly reasons about uncertainty through strong spatial priors. Using a new behavioral Map Induction Task, we demonstrate that this computational framework explains human exploration behavior better than non-inductive models and outperforms state-of-the-art planning algorithms in a realistic spatial navigation domain.

Efficient exploration of spatial environments through Map Induction using adaptable compositional map representations

Inhibition-of-return (IOR) is demonstrated by longer response times (RTs) for stimuli presented at previously attended locations than unattended locations. While facilitatory cueing effects occur at short stimulus onset asynchronies (SOAs) between a cue and target, IOR emerges at later SOAs. Frontoparietal areas including posterior parietal cortex have been implicated in producing IOR, but the contribution of primary motor cortex (M1) is unclear. 
In the present study, participants executed key-press responses to peripheral targets following a cue presented at the same or opposite location at different SOAs (100/300/600/1000ms). Single-pulse transcranial magnetic stimulation (TMS) was applied over right M1 on 50% of trials in a randomized (Experiment 1) or blocked (Experiment 2) order.
IOR appeared to be attenuated by TMS, although RT profiles differed between experiments. Motor-evoked potentials were not altered by the cue-target relationship in either experiment. These findings do not support a key role of M1 in mechanisms of IOR.

The role of primary motor cortex in inhibition of return: An investigation using transcranial magnetic stimulation

Navigation is a fundamental cognitive function of virtually all
moving animals. Several navigation strategies require an
estimate of the current travelling direction that is updated
continuously. In the central complex of the insect brain,
multimodal cues are fused into a compass-like head direction
representation. Based on the proposed connectivity of
columnar neurons in the central complex of the desert locust
we designed a computational model to examine how these
neurons could maintain a stable representation of heading
direction and how shifts occur by optic flow signals when the
animal turns. Our proposed model architecture shows that the
activity of head direction-encoding CL1a neurons remains
stable if the activity of a second class of columnar neurons,
CL2, is exactly the same. Shifts occur via modulation of the
network connectivity. Our model can be used to deduce
testable hypotheses where data are lacking, inspiring new
avenues of experimental investigations.

A Model for Optic Flow Integration in Locust Central-Complex Neurons Tuned to Head Direction

Speech errors are often perceived as categorical substitutions of one sound for another, but phonetic analyses have consistently revealed that errorful productions retain a phonetic trace of the target category. These trace effects have been taken as evidence for the simultaneous activation of multiple categories, both exerting influence on speech production. We develop a dynamic neural field model of voice onset time (VOT) planning, showing how multiple activated categories can be resolved in the field to show trace effects. We evaluate model predictions against measurements of VOT for voiced and voiceless stops in speech error experiments and naturalistic corpora.


A dynamic neural field model of phonetic trace effects in speech errors

In the debate between single-route and dual-route models of verb inflection, the dissociation between regular and irregular verbs in the non-fluent variety of aphasia has been a key sticking point for the proponents of the single-route model. This paper adopts a state-of-the-art neural model which has previously been used to learn inflectional morphology, and shows that it can also be used to model data from non-fluent aphasia. This challenges the assumption that a dual-route model is necessary to capture apparent dissociations in aphasia data and encourages a reanalysis of the deficits involved in non-fluent aphasia.


Premium content

Hippocampus Engagement Early in Learning Supports Concept Formation

Next from CogSci 2022

Identifying a Phonetic Factors of Onomatopoeias Correlated to Sound Symbolic Commons between Japanese and Non-Japanese Speakers

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES