Overview

• The adaptive computation framework describes how limited perceptual resources are allocated to prioritize goal-relevant information in dynamic scenes.
• Researchers conducted computer-based experiments in which volunteers tracked highlighted circles moving among identical distractors to test the model.
• Model predictions aligned with participants’ sub-second focus changes and subjective assessments of task difficulty across varying distraction levels.
• Data revealed a computational signature of cognitive effort, linking the model’s resource allocation to perceived exertion during prolonged attention tasks.
• Yale team members Ilker Yildirim and Mario Belledonne published their findings in Psychological Review on June 26 and 27, highlighting potential applications for more human-like AI systems.