ICCM: ACT-R

Temporal binding (TB) is the subjective compression between a voluntary action and its associated outcome. It is regarded as an implicit measure of the sense of agency; however, an underlying mechanism has yet to be agreed upon. Previous research suggests memory as an alternative explanation for TB in two publicly available datasets. Here, we test this idea by implementing a model within the ACT-R cognitive architecture and leveraging its existing memory and time perception mechanisms to simulate participants from these datasets. Our model simulations provide evidence to suggest that memory and time perception mechanisms can explain the pattern of results. Implications for temporal binding and the sense of agency will be discussed.

This is an in-person presentation on July 21, 2023 (09:00 ~ 09:20 UTC).

The environment surrounding organisms changes dynamically, and humans acquire motor skills by improving the prediction of such environmental changes. The research on cognitive architectures has so far proposed several mechanisms explaining the process of human motor learning. Adaptive Control of Thought-Rational (ACT-R), one of the representative cognitive architectures, has perceptual and motor modules for interaction with the external environment. However, the performance of these modules is insufficient for real-time environments, especially in terms of learning speed. This study proposes a method to simulate human-level rapid motor learning using a pre-trained motor learning module. We assume that in a novel perceptual-motor task, a pre-trained motor schema is rediscovered/recalled. In the simulations, we trained the motor learning module in advance and conducted a simulation where difficulties of rediscovering schemata were manipulated. As a result, we confirmed that the pre-trained phase increased the human-model fitting in motor learning.

This is an in-person presentation on July 21, 2023 (09:20 ~ 09:40 UTC).

In adversarial operational environments like radar monitoring, humans have to monitor large amounts of information, multitask, and manage threats. They may also face electronic disruption or attacks aimed at degrading radar monitor effectiveness (a.k.a electronic warfare or EW). In these settings, it is unclear how frequent changes in personnel, training, and updates to visual displays affect an operator's readiness. A recent experiment used an analogous radar monitoring task to investigate effects of display density and electronic warfare on an operator's threat detection performance. Here, we present a cognitive model capable of completing a scaled down version of that task to better understand the experimental results and underlying cognitive processes. Similar to the human experiment, our cognitive model completed conditions comprised of changes to the nature of the task(s), the number of targets to track, and the presence or absence of distractors, deemed 'friendlies'. Although this initial cognitive model uses primarily default ACT-R parameters, it was able to capture patterns in human performance across conditions. We present the results and discuss limitations to address in future work.

This is an in-person presentation on July 21, 2023 (09:40 ~ 10:00 UTC).

Symbolic/hybrid computational cognitive architectures, including the ACT-R framework, are adept at capturing a wide variety of human cognitive processes and behaviors including problem-solving, memory, and language. However, such cognitive architectures do not capture visuomotor behaviors that tightly couple perceptual and motor processes – such as manual tracking. In this study, we aimed to improve the cognitive fidelity of manual tracking behavior within the ACT-R framework by implementing the position control model (PCM) – a continuous, linear control model that effectively captures human tracking behavior (Powers, 1978). We integrated PCM within a MATB task model developed within the ACT-R framework, to examine if the integrated ACT-R/PCM model showed improvement in capturing human tracking performance relative to the Standard ACT-R model. Results indicate that the ACT-R/PCM Integrated model showed improved performance in capturing certain aspects of human tracking behavior, in comparison to the Standard ACT-R model.

This is an in-person presentation on July 21, 2023 (10:00 ~ 10:20 UTC).

We present a cognitive model that plays a video game of driving a bus for a long time. The model was built using the ACT-R cognitive architecture and an extension to support perceptual-motor knowledge of how to interact with the environment (VisiTor and ACT-R/PM). Our extension includes bitmap-level eyes and robot hands. The model was run for a long time, over 6 hours on the way from Tucson to Las Vegas. We employed a design approach based on the ADDIE model to create different knowledge representations and actions; the model’s predictions can be matched to some aspects of human behavior on the fine details regarding the number of course corrections and average speed and learning rate. However, it does not exhibit the same level of fatigue as human behavior. This contrasts with the way humans typically perform such long tasks. This model shows that perception opens up new interfaces and provides a very accessible testbed for examing further aspects of behavior. and adding components of human behavior that remain missing from ACT-R.

This is an in-person presentation on July 21, 2023 (10:20 ~ 10:40 UTC).