Faster but less precise: expectation enhances response speed while reducing sensory fidelity

The brain's remarkable ability to process continuous sensory inputs with adaptive efficiency - balancing flexibility while minimizing metabolic cost - is thought to rely on predictive mechanisms that generate and update internal models that leverage statistical regularities in the environment. However, it remains unclear whether this efficiency arises from prioritizing reliable, expected events or informative, unexpected ones, as they offer complementary adaptive advantages. To isolate genuine expectation effects, we combined electroencephalography (EEG), pupillometry, and behavioural measures in a paradigm that independently manipulated task relevance (selective attention) and stimulus predictability, while minimizing stimulus repetition at identical spatial locations to control for low-level adaptation. Human participants (both sexes) responded faster and more accurately to expected events, which was enhanced when attention was engaged; however, these events were reproduced with lower precision, independent of attention. Feature-specific neural decoding revealed pre-stimulus effects of attention and post-stimulus effects of expectation, with no interaction between the two. Attention increased decoding accuracy, while expectation reduced accuracy. The reduced representational fidelity for expected events appeared rapidly (∼100-200 ms after stimulus onset) and correlated with individual differences in perceptual precision. Collectively, our findings indicate two complementary processes that define how the brain leverages redundancy in the environment: an early (pre-stimulus) mechanism, which supports rapid motor responses to expected events and is mediated by attention, and a later (post-stimulus) process, which dampens sensory responses to expected events and is unaffected by attention.Significance statement The brain must process vast amounts of sensory information efficiently while remaining flexible enough to detect important changes. Competing theories propose that the brain saves energy either by prioritizing expected events or by enhancing responses to surprising ones, but experimental evidence has been inconsistent. Our findings show that the brain uses both strategies, each serving a different purpose. Expected events are processed quickly to support rapid, accurate actions, whereas expected events are also encoded with reduced precision, prioritising unexpected events to update internal models and guide future behaviour. By cleanly separating expectation from attention and low-level adaptation, we provide a unified explanation of how the brain balances efficiency with adaptability across multiple timescales.