A drift-diffusion model of temporal generalization outperforms existing models and captures modality differences and...

Area of Science:

• Cognitive Neuroscience
• Computational Neuroscience
• Decision Making

Background:

• The brain employs multiple systems for time perception, adapting to temporal demands.
• Neural timing mechanisms vary across brain regions and tasks.
• Drift-diffusion models offer a computational framework for understanding timing behaviors.

Purpose of the Study:

• To develop and validate a drift-diffusion model for the temporal generalization task.
• To investigate how sensory modality (vision vs. audition) and learning affect temporal decision-making parameters.

Main Methods:

• Developed a novel drift-diffusion model incorporating an internal time estimate and decision boundaries.
• Applied the model to fit participant data from temporal generalization experiments.
• Analyzed the impact of sensory modality and learning on model parameters.

Main Results:

• The proposed drift-diffusion model demonstrated superior fit to participant data compared to previous models.
• Decision boundaries were found to be independently modifiable: higher in vision than audition, and decreasing with learning.
• Timing noise positively correlated with upper decision boundaries, suggesting an accuracy-maximizing strategy.

Conclusions:

• The drift-diffusion model provides a robust framework for understanding temporal generalization and decision-making.
• Sensory modality and learning dynamically adjust decision boundaries to optimize performance.
• Neural timing strategies appear to maximize accuracy in temporal tasks.