Learning to perform role-filler binding with schematic knowledge

Catherine Chen; Qihong Lu; Andre Beukers; Christopher Baldassano; Kenneth A Norman

doi:10.7717/peerj.11046

Learning to perform role-filler binding with schematic knowledge

PeerJ. 2021 Mar 31:9:e11046. doi: 10.7717/peerj.11046. eCollection 2021.

Authors

Catherine Chen¹, Qihong Lu², Andre Beukers², Christopher Baldassano³, Kenneth A Norman²

Affiliations

¹ Department of Computer Science, Princeton University, Princeton, NJ, USA.
² Department of Psychology, Princeton University, Princeton, NJ, USA.
³ Department of Psychology, Columbia University, New York, NY, USA.

Abstract

Through specific experiences, humans learn the relationships that underlie the structure of events in the world. Schema theory suggests that we organize this information in mental frameworks called "schemata," which represent our knowledge of the structure of the world. Generalizing knowledge of structural relationships to new situations requires role-filler binding, the ability to associate specific "fillers" with abstract "roles." For instance, when we hear the sentence Alice ordered a tea from Bob, the role-filler bindings customer:Alice, drink:tea and barista:Bob allow us to understand and make inferences about the sentence. We can perform these bindings for arbitrary fillers-we understand this sentence even if we have never heard the names Alice, tea, or Bob before. In this work, we define a model as capable of performing role-filler binding if it can recall arbitrary fillers corresponding to a specified role, even when these pairings violate correlations seen during training. Previous work found that models can learn this ability when explicitly told what the roles and fillers are, or when given fillers seen during training. We show that networks with external memory learn to bind roles to arbitrary fillers, without explicitly labeled role-filler pairs. We further show that they can perform these bindings on role-filler pairs that violate correlations seen during training, while retaining knowledge of training correlations. We apply analyses inspired by neural decoding to interpret what the networks have learned.

Keywords: Fillers; Frames; Neural networks; Role-filler binding; Roles; Schema; Schemata.

Grants and funding

Funding was provided by a grant from Intel Labs and by a Multi-University Research Initiative grant to Kenneth A. Norman (ONR/DoD N00014-17-1-2961). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.