Intrinsic Motivation to Learn Action-State Representation with Hierarchical Temporal Memory

Dzhivelikian, Evgenii; Latyshev, Artem; Kuderov, Petr; Panov, Aleksandr I.

doi:10.1007/978-3-030-86993-9_2

Cited by 3 publications

(2 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This work builds on preliminary findings presented in the conference paper [ 22 ]. This paper provides a more detailed description of our hierarchical memory model than that in the conference paper.…”

Section: Introductionmentioning

confidence: 67%

See 1 more Smart Citation

Hierarchical intrinsically motivated agent planning behavior with dreaming in grid environments

et al. 2022

Self Cite

View full text Add to dashboard Cite

Biologically plausible models of learning may provide a crucial insight for building autonomous intelligent agents capable of performing a wide range of tasks. In this work, we propose a hierarchical model of an agent operating in an unfamiliar environment driven by a reinforcement signal. We use temporal memory to learn sparse distributed representation of state–actions and the basal ganglia model to learn effective action policy on different levels of abstraction. The learned model of the environment is utilized to generate an intrinsic motivation signal, which drives the agent in the absence of the extrinsic signal, and through acting in imagination, which we call dreaming. We demonstrate that the proposed architecture enables an agent to effectively reach goals in grid environments.

show abstract

Section: Introductionmentioning

confidence: 67%

“…There are works extending its usage in part [ 21 , 30 , 38 ]. However, all of these issues were first to be addressed using an original approach by Dzhivelikian et al [ 22 ]. Our current work is its direct extension, in which we contribute to the analysis of spatial–temporal abstractions arisen in such hierarchy.…”

Section: Related Workmentioning

confidence: 99%