Perception is a process of inference, integrating sensory inputs with prior expectations. However, little is known regarding the temporal dynamics of this integration. It has been proposed that expectation plays a role early in the perceptual process, biasing sensory processing. Alternatively, others suggest that expectations are integrated only at later, postperceptual decision-making stages. The current study aimed to dissociate between these hypotheses. We exposed human participants (male and female) to auditory cues predicting the likely direction of upcoming moving dot patterns, while recording neural activity using magnetoencephalography (MEG). Participants' reports of the moving dot directions were biased toward the direction predicted by the cues. To investigate when expectations affected sensory representations, we used inverted encoding models to decode the direction represented in early sensory signals. Strikingly, the cues modulated the direction represented in the MEG signal as early as 150 ms after visual stimulus onset. While this may not reflect a modulation of the initial feedforward sweep, it does reveal a modulation of early sensory representations. Exploratory analyses showed that the neural modulation was related to perceptual expectation effects: participants with a stronger perceptual bias toward the predicted direction also revealed a stronger reflection of the predicted direction in the MEG signal. For participants with this perceptual bias, a correlation between decoded and perceived direction already emerged before visual stimulus onset, suggesting that the prestimulus state of the visual cortex influences sensory processing. Together, these results suggest that expectations play an integral role in the neural computations underlying perception.

Considerable evidence supports the utility of the concepts of cognitive fusion and defusion, across diverse areas of functioning and concern. Cognitive fusion includes taking thoughts literally, and thinking being likely to dominate behaviour. Acceptance and Commitment Therapy (ACT) protocols include 'defusion' exercises, to support a distanced and more flexible relationship with thoughts. The Cognitive Fusion Questionnaire (CFQ), is a brief, psychometrically sound measure of the process. Although the CFQ has generated fusionfocussed research, it does not tend to be used in experimental settings. Instead, fusionfocussed experiments tend to use psychometrically untested measures of thought believability as a proxy for fusion. This paper reports on the initial validation of a state version of the CFQ, the State Cognitive Fusion Questionnaire (SCFQ). Three separate, predominantly student-based convenience samples, with a combined n of 379 provided preliminary evidence that the SCFQ has a theoretically consistent factor structure, shows excellent internal reliability, is stable over time, and is a valid measure of fusion. The findings also demonstrate that the SCFQ is more sensitive than the CFQ to changes in fusion, following a brief defusion exercise, making it more suitable for lab based studies.

No abstract