Matthew Botvinick

TL;DR: Perceiver AR is developed, an modality-agnostic architecture which uses cross-attention to map long-range inputs to a small number of latents while also maintaining end-to-end causal masking, enabling practical long-context density estimation without the need for hand-crafted sparsity patterns or memory mechanisms.

TL;DR: This chapter reviews two results that show the existence of neural correlates to key predictions from HRL, and focuses on one aspect of this work, which deals with the question of how action hierarchies are initially established.

TL;DR: A computational model of object and action selection is developed and analyzed, established through a series of simulations, that the impact of distractor objects on reaching times can vary depending on the nature of the current action plan.

TL;DR: In this paper , a deep learning system was proposed to learn intuitive physics directly from visual data, inspired by studies of visual cognition in children, which can learn a diverse set of physical concepts, which depends critically on object-level representations, consistent with findings from developmental psychology.

TL;DR: This model's implications for understanding task representation are considered, considering the implications of the account for two influential concepts: (1) cognitive underspecification, the idea that task representations may be imprecise or vague, especially in contexts where errors occur, and (2) information-sharing,The idea that closely related operations rely on common sets of internal representations.