Related Experiment Video
Updated: Apr 12, 2026
Immunohistochemical Visualization of Hippocampal Neuron Activity After Spatial Learning in a Mouse Model of Neurodevelopmental Disorders
Published on: May 12, 2015
Human hippocampal ripples prioritize model-based learning
Xiaoyu Zhou1, Xiongfei Wang2, Xiangyu Hu1
1State Key Laboratory of Cognitive Neuroscience and Learning, IDG/McGovern Institute for Brain Research, Beijing Normal University, Beijing 100875, China; Chinese Institute for Brain Research, Beijing 102206, China.
The brain uses hippocampal ripples and prefrontal cortex coordination to support model-based learning, enabling efficient inference of unexperienced option values. This neural mechanism prioritizes updating valuable indirect experiences for better decision-making.
Area of Science:
- Neuroscience
- Cognitive Science
- Computational Neuroscience
Background:
- Humans demonstrate advanced model-based learning, utilizing internal world models for inferring unexperienced option values from sparse data.
- The neural underpinnings of this sophisticated learning mechanism, particularly how the brain supports inferring values of unvisited paths, remain largely unexplored.
Purpose of the Study:
- To investigate the neural mechanisms supporting model-based reinforcement learning, specifically focusing on the role of the hippocampus and prefrontal cortex in inferring values of unvisited options.
- To examine the relationship between hippocampal activity, cortical reactivation, and prefrontal cortex engagement during a task requiring inference of non-local path values.
Main Methods:
- Intracranial electrophysiology (iEEG) was recorded from 34 epilepsy patients during a reinforcement-learning task.
- The task design necessitated inferring values of unvisited (non-local) paths using task structure.
- Analysis focused on hippocampal ripples, cortical reactivation patterns, and lateral frontopolar cortex activity in relation to learning outcomes.
Main Results:
- Hippocampal ripples were significantly associated with prioritized learning about non-local paths.
- Ripple events post-reward contained information about the most valuable indirect experience to update, with longer ripples indicating stronger priority signals.
- Activity in the lateral frontopolar cortex synchronized with hippocampal ripples, predicting enhanced use of task structure and more accurate non-local value learning.
Conclusions:
- Hippocampal ripples play a crucial role in prioritizing information for model-based learning, particularly for updating values of unexperienced options.
- Ripple-centered coordination between the hippocampus and the lateral frontopolar cortex is essential for efficient model-based reinforcement learning.
- These findings elucidate a key neural circuit supporting sophisticated cognitive functions like model-based inference and decision-making.
