English
 
User Manual Privacy Policy Disclaimer Contact us
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT
  Forgetful inference in a sophisticated world model

Ahilan, S., Solomon, R., Conover, K., Niyogi, R., Shizgal, P., & Dayan, P. (2017). Forgetful inference in a sophisticated world model. Poster presented at Conference on Cognitive Computational Neuroscience (CCN 2017), New York, NY, USA. doi:10.1101/419317.

Item is

Basic

show hide
Item Permalink: http://hdl.handle.net/21.11116/0000-0004-8980-0 Version Permalink: http://hdl.handle.net/21.11116/0000-0004-B96A-5
Genre: Poster

Files

show Files

Creators

show
hide
 Creators:
Ahilan, S, Author
Solomon, R, Author
Conover, K, Author
Niyogi, R, Author
Shizgal, P, Author
Dayan, P1, Author              
Affiliations:
1External Organizations, ou_persistent22              

Content

show
hide
Free keywords: -
 Abstract: Humans and other animals are able to discover underlying statistical structure in their environments and exploit it to achieve efficient and effective performance. However, the largest scale structures such as 'world models' are often difficult to learn and use because they are obscure, involving long-range temporal dependencies. We analyzed behavioral data from an extended experiment with rats, showing that the subjects discovered and exploited a world model, albeit suffering at times from immediate inferential imperfections as to their current state within it. To describe this process, we built a hidden Markov model (HMM) of the subjects' models of the experiment, describing overall behavior as integrating recent observations with the recollections of an imperfect memory. Over the course of training, we found that subjects came to track their progress through the task more accurately, indicating improved inference of the partially-observable state. Model fits attributed this improvement to decreased forgetting of the previous state. This `learning to remember' decreased reliance on more recent observations, which can be misleading, in favor of a more dependable memory.

Details

show
hide
Language(s):
 Dates: 2017-09
 Publication Status: Published online
 Pages: -
 Publishing info: -
 Table of Contents: -
 Rev. Method: -
 Identifiers: DOI: 10.1101/419317
 Degree: -

Event

show
hide
Title: Conference on Cognitive Computational Neuroscience (CCN 2017)
Place of Event: New York, NY, USA
Start-/End Date: 2017-09-06 - 2017-09-08

Legal Case

show

Project information

show

Source 1

show
hide
Title: Conference on Cognitive Computational Neuroscience (CCN 2017)
Source Genre: Proceedings
 Creator(s):
Affiliations:
Publ. Info: -
Pages: - Volume / Issue: - Sequence Number: 3000034 Start / End Page: - Identifier: -