<p>Retrospective cues are thought to prioritize goal-relevant representations in visual working memory, yet whether they actively suppress interference from reward-associated non-targets remains unclear. We combined a reward-association training phase with a retro-cue paradigm across two eye-tracking experiments to test how retrospective attention regulates value-driven distraction within memory. When non-reward stimuli were absent during training (Experiment 1), high-reward items produced smaller adjustment errors, lower guess rates, and higher precision than low-reward items, with no reliable target–non-target difference, indicating strong value-based competition. When non-reward stimuli were included during training (Experiment 2), high–low reward differences disappeared, whereas targets outperformed non-targets on error and precision, consistent with retro-cue–based selection. Microsaccade analyses showed that, irrespective of training, post-cue gaze dynamics favored targets. Critically, in Experiment 1, when reward stimuli served as non-targets, microsaccades were less frequently directed toward high-reward than low-reward distractors, providing direct oculomotor evidence for suppression of value-laden interference; this difference was absent in Experiment 2. Together, these results indicate that retrospective attention suppresses reward-associated interference within working memory, and that this suppression is gated by stimulus novelty introduced during training. The findings link behavioral precision and oculomotor dynamics to an active suppression mechanism that controls value-driven distraction after encoding, offering a principled account of how internal attention resolves competition among mnemonic representations.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Retrospective attention suppresses reward-associated interference in visual working memory

  • Yaya Zhang,
  • Shuya Lei,
  • Gongao Li,
  • Xuezhu Hu,
  • Jinhong Ding

摘要

Retrospective cues are thought to prioritize goal-relevant representations in visual working memory, yet whether they actively suppress interference from reward-associated non-targets remains unclear. We combined a reward-association training phase with a retro-cue paradigm across two eye-tracking experiments to test how retrospective attention regulates value-driven distraction within memory. When non-reward stimuli were absent during training (Experiment 1), high-reward items produced smaller adjustment errors, lower guess rates, and higher precision than low-reward items, with no reliable target–non-target difference, indicating strong value-based competition. When non-reward stimuli were included during training (Experiment 2), high–low reward differences disappeared, whereas targets outperformed non-targets on error and precision, consistent with retro-cue–based selection. Microsaccade analyses showed that, irrespective of training, post-cue gaze dynamics favored targets. Critically, in Experiment 1, when reward stimuli served as non-targets, microsaccades were less frequently directed toward high-reward than low-reward distractors, providing direct oculomotor evidence for suppression of value-laden interference; this difference was absent in Experiment 2. Together, these results indicate that retrospective attention suppresses reward-associated interference within working memory, and that this suppression is gated by stimulus novelty introduced during training. The findings link behavioral precision and oculomotor dynamics to an active suppression mechanism that controls value-driven distraction after encoding, offering a principled account of how internal attention resolves competition among mnemonic representations.