Are short-term memories just noisier versions of what we perceive? Are they fundamentally different? We (Chaipat Chunharas, @mjwolff.bsky.social, @meikehettwer.bsky.social and myself) delved into this in a paper out now in #elife: elifesciences.org/reviewed-pre.... For a quick summary, a 🧵 below: - ThreadSky

rademaker.bsky.social • 56 days ago

Are short-term memories just noisier versions of what we perceive? Are they fundamentally different? We (Chaipat Chunharas, @mjwolff.bsky.social, @meikehettwer.bsky.social and myself) delved into this in a paper out now in #elife: https://elifesciences.org/reviewed-preprints/103347. For a quick summary, a 🧵 below:

Comments

achetverikov.bsky.social•56 days ago

Very cool work! I wonder if it is possible to distinguish more categorical representations from more biased ones? And what about uncertainty of the representations? Finally, how does decoding correspond to behavior? Decoders might have their own biases not necessarily reflecting behavior.

rademaker.bsky.social•55 days ago

Tnx Andrey! Repulsion biases (from cardinal) might be a form of categorization actually? Or rather, a behavioral expression of it. I looked at uncertainty but not with tafkap so I wouldn’t dare make claims ;) And wrt behavior - there kind of wasn’t any for the perceptual task (contrast detection)

achetverikov.bsky.social•55 days ago

Yes, exactly, I was wondering whether there is a conceivable difference here between categorization and biases away from cardinals. I mean, uncertainty would probably help to distinguish them, but I'm not so sure.

rademaker.bsky.social•56 days ago

Early visual cortex processes what we see around us, but also has information about images briefly held in mind. The two must be different... But different how?

rademaker.bsky.social•56 days ago

To answer how perception and memory differ in visual cortex, we use line-images ("gratings") with different orientations that people look at, or remember, while we record voxel responses. We correlate patterns of responses evoked by all possible orientations & VOILA, "representational geometry"

rademaker.bsky.social•56 days ago

Whoa... Those don't look the same at all! The geometry during perception is more like a diagonal line, and during short-term memory it's more like... 2 squares? There's also a change from earlier (V1) to later (IPS) visual areas...

rademaker.bsky.social•56 days ago

So what's going on? As it happens, Chaipat & I have often wondered if things in memory are perhaps more categorical, while the things we perceive are more true to reality (i.e., "veridical"). So let's make some models!

rademaker.bsky.social•56 days ago

Girshick, Landy, & Simoncelli (2011) showed it best: The natural world has more horizontal and vertical ("cardinal") orientations in it than those in between ("obliques"). We start from this uneven probability distribution over orientation.

rademaker.bsky.social•56 days ago

We model what similarity of fMRI responses might look like when assuming that (1) the brain uses a veridical way to represent orientations, or (2) the brain categorizes orientations based on how different they appear in a more experiential sense.

Comments

Posting Rules

Reply