English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Animal navigation research posits that organisms build and maintain internalspatial representations, or maps, of their environment. We ask if machines --specifically, artificial intelligence (AI) navigation agents -- also buildimplicit (or 'mental') maps. A positive answer to this question would (a)explain the surprising phenomenon in recent literature of ostensibly map-freeneural-networks achieving strong performance, and (b) strengthen the evidenceof mapping as a fundamental mechanism for navigation by intelligent embodiedagents, whether they be biological or artificial. Unlike animal navigation, wecan judiciously design the agent's perceptual system and control the learningparadigm to nullify alternative navigation mechanisms. Specifically, we train'blind' agents -- with sensing limited to only egomotion and no other sensingof any kind -- to perform PointGoal navigation ('go to $\Delta$ x, $\Delta$ y')via reinforcement learning. Our agents are composed of navigation-agnosticcomponents (fully-connected and recurrent neural networks), and ourexperimental setup provides no inductive bias towards mapping. Despite theseharsh conditions, we find that blind agents are (1) surprisingly effectivenavigators in new environments (~95% success); (2) they utilize memory overlong horizons (remembering ~1,000 steps of past experience in an episode); (3)this memory enables them to exhibit intelligent behavior (following walls,detecting collisions, taking shortcuts); (4) there is emergence of maps andcollision detection neurons in the representations of the environment built bya blind agent as it navigates; and (5) the emergent maps are selective and taskdependent (e.g. the agent 'forgets' exploratory detours). Overall, this paperpresents no new techniques for the AI audience, but a surprising finding, aninsight, and an explanation.

Emergence of Maps in the Memories of Blind Navigation Agents