"Nothing" Really Matters: What Omission Responses Reveal About the Predictive Brain

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Understanding endogenous brain activity is crucial for deciphering predictive mechanisms.
• Omission paradigms offer a unique method to study internal brain computations by examining responses to absent expected stimuli.

Purpose of the Study:

• To synthesize existing research on omission paradigms across species, modalities, and experimental designs.
• To reveal how the brain generates predictions and anticipates the world.
• To propose a unified framework for understanding omission responses and their underlying neural computations.

Main Methods:

• Comprehensive literature review of omission research.
• Analysis of empirical signatures such as timing, attention dependence, and content-specificity.
• Development of a functional "Omission Atlas" mapping computations to brain networks.

Main Results:

• The brain utilizes explicit learned models to create detailed representations of expected stimuli.
• Anticipatory signals and "sensory ghosts" of absent stimuli provide evidence for predictive processing.
• Omission responses arise from two cooperative computational styles: Local Regularity Encoding (LRE) and Model-Based Inference (MBI).

Conclusions:

• A unified framework of LRE and MBI explains omission responses, from rapid local computations to flexible, attention-dependent predictions.
• The "Omission Atlas" maps these computations across diverse brain networks.
• This framework advances understanding of predictive processing and has implications for neurological and psychiatric disorders.