Post

@williamngiam.github.io

Worth reading the developing thread below! I've been thinking along similar lines, but instead of more labels, I have been thinking moreso in a mathematical modelling framework: what we consider a "representation" and what we consider a "process" (partly inspired by Grace's book Models of the Mind!)

This post is unavailable.

14/10/26, 12:19 pm0 reposts3 likes

William Ngiam | 严祥全

@williamngiam.github.io

I think many conceptualise the different fields of visual cognition (attention, working memory, etc.) as different mechanisms/processes but without definition of what those mechanisms are operating upon (i.e. the "representation"). What if it's the same process, but simply different representations?

William Ngiam | 严祥全

@williamngiam.github.io

I think about all the models (evidence accumulation, signal-detection, etc.) that can be applied to a variety of contexts. Are we all studying the same process – a Thurstonian model of latent variables in the mind – but simply with different contexts? Or are there different mechanisms we can detail?

William Ngiam | 严祥全

@williamngiam.github.io

Anyway, it's not straightforward to distinguish what is a representation and what is a process; perhaps it does not matter. We (@mdlbayes.bsky.social and I) argue for joint models of both (defining each as best as we can, and be explicit about what we are assuming); paper coming soon hopefully...

osf.ioOSF