As within, so without, as above, so below: Common mechanisms can support between- and within-trial category learning...

Area of Science:

• Cognitive Psychology
• Computational Neuroscience
• Machine Learning

Background:

• Category learning involves challenges in identifying relevant information and its optimal timing.
• Existing theories inadequately address within-trial attention allocation and temporal information sampling during learning.

Purpose of the Study:

• To introduce and validate the adaptive attention representation model (AARM) for category learning.
• To elucidate how attention distribution updates between trials and shifts dynamically within trials.
• To investigate the interplay between attention, decision dynamics, and gaze behavior during category learning.

Main Methods:

• Development and application of the adaptive attention representation model (AARM).
• Validation against behavioral data from four distinct case studies on selective attention.
• Utilizing eye-tracking and choice response data to analyze dynamic interactions between attention and decision processes.

Main Results:

• AARM successfully models the dynamic updating and shifting of attention during category learning.
• The model accounts for how attention allocation influences decision dynamics.
• Empirical data supports AARM's predictions regarding the interaction of attention and decision-making.

Conclusions:

• A common set of mechanisms can explain both between-trial attention updates and within-trial attention shifts.
• The adaptive attention representation model provides a unified framework for understanding attention dynamics in category learning.
• Findings highlight the bidirectional influence between attentional processes and decision dynamics in learning.