Definitely written with AI, as well as the other two papers making the same point. Extensive use of 'em dashes', AI generated figures with 3 finger hands and all. Bit dodgy.

Symposium on setbacks in science! @imagingneuroucl.bsky.social @roelandheerema.bsky.social Looking forward to share my rocky journey through science with everyone

Ladies and Gentlemen, I can officially say I've made it!

*tap* *tap* Is this thing still on? After leaving Xitter last week, I will exclusively post my research updates on Bsky. But first, one more month of paternity leave!

We found that both on veridical and false percept trials that there was stimulus-like and orientations-specific activity representing what participants saw. In the case of false percepts there was additional pre-stimulus activity.

We then decoded participants confidence in stimulus presence on both absent and present trials, and found a confidence signal emerging 250-300ms post-stimulus emerging from a parietal frontal network.

These high confidence false percepts were preceded by a relative increase in oscillatory beta power originating from a parietal occipital network, which was not present on grating present trials. Beta power increased after the onset of the predictive auditory cue.

As in our previous study, participants orientation responses were only driven by predictive auditory cues if they were aware of them. But just like in our previous study, participants indicated on a third of noise-only trials that they were confident that there was a grating.

Finally, separate online study, explored whether behaviour on this task correlated with hallucination-proneness. We found that confidence in false alarms as well as reduced sensory precision correlated with hallucinations (also recently found by Benrimoh tinyurl.com/yubc9hmt).

In contrast, the orientation that was cued was represented in the deep-layers, replicating our previous work (Aitken, 2020), and subsequently replicated again in a study in collaboration with
Emily Thomas
&
Clare Press
(2023). tinyurl.com/nprsw5yb tinyurl.com/45rvyfyj

We found that the orientation of these high confidence false alarms were represented by the middle layers of the early visual cortex, suggesting, in line with behaviour, that they might arise through feedforward mechanisms.

Behaviourly, participants were not influenced by the auditory cues, unless they had become aware of what orientation these cues predicted. Still, participants reported seeing a grating with high levels of confidence on absent trials. What drove these reports if not the cues?

25 participants performed a difficult perceptual discrimination task under noisy conditions while undergoing a laminar fMRI scan at 7T. Participants were presented with either clockwise or anticlockwise gratings on 50% of the trials, or just noise on the remaining 50%.