iiinova — where your digital self keeps evolving

Replies

William Ngiam | 严祥全 @williamngiam.github.io · 13/07/2026

The reviewers pointed out the joint modelling approach means that the inferred representation depends on the adequacy of the assumed processes in our models. At the least, in supp material, we show the task-specific models provide adequate explanations of the data: github.com/mdlee/orient... /3

github.com

GitHub - mdlee/orientationModeling

Contribute to mdlee/orientationModeling development by creating an account on GitHub.

110

William Ngiam | 严祥全 @williamngiam.github.io · 13/07/2026

Our work scrutinises whether we can assume a common (similarity-based) representation for cognition, but we agree with the reviewers' critique that improvement of our models can change the inference. We discuss two future approaches to this: the unification approach and the cataloging approach. /4

110

William Ngiam | 严祥全 @williamngiam.github.io · 13/07/2026

The unification approach tries to combine the task-specific models and resolve a single underlying representation. This can be achieved by adding principled processes explaining variation across tasks (e.g. selective attention in Nosofsky's work) or by altering how the representation is defined. /5

110