A programmable neural virtual machine based on a fast store-erase learning rule

Katz, Garrett E.; Davis, Gregory P.; Gentili, Rodolphe J.; Reggia, James A.

doi:10.1016/j.neunet.2019.07.017

Cited by 8 publications

(9 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It has been shown that these recall and storage rules perfectly replicate the semantics of a symbolic key-value lookup table, as long as certain conditions are met (Katz et al, 2019 ). Specifically, distinct key symbols must be represented by orthogonal pattern vectors with ±1 elements.…”

Section: Introductionmentioning

confidence: 99%

“…Programmable neural networks (Verona et al, 1991 ; Neto et al, 2003 ; Eliasmith and Stewart, 2011 ; Bošnjak et al, 2017 ; Katz et al, 2019 ; Davis et al, 2021 ) comprise one potential approach to building such systems. These are neural networks whose dynamics can emulate execution of human-authored source code.…”

Section: Introductionmentioning

confidence: 99%

“…In this paper, we test this approach using a programmable neural network model called the “Neural Virtual Machine” (NVM), which is asymptotically Turing-complete as layer size grows (Katz et al, 2019 ). We program the NVM to control the Poppy™ Ergo Jr robotic manipulator (Lapeyre et al, 2014 ) in a PyBullet (Coumans and Bai, 2021 ) simulation environment.…”

Section: Introductionmentioning

confidence: 99%

“…Our programmable neural network used here is an updated version of our recently proposed “Neural Virtual Machine” (NVM) architecture (Katz et al, 2019 ). The following section provides the requisite background information needed to understand the NVM and how we use it here.…”

Section: Introductionmentioning

confidence: 99%

“…The performed (ungated) operation and the layers involved are shown in black. Adapted from Katz et al ( 2019 ). (A) Increment.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Tunable Neural Encoding of a Symbolic Robotic Manipulation Algorithm

et al. 2021

Self Cite

View full text Add to dashboard Cite

We present a neurocomputational controller for robotic manipulation based on the recently developed “neural virtual machine” (NVM). The NVM is a purely neural recurrent architecture that emulates a Turing-complete, purely symbolic virtual machine. We program the NVM with a symbolic algorithm that solves blocks-world restacking problems, and execute it in a robotic simulation environment. Our results show that the NVM-based controller can faithfully replicate the execution traces and performance levels of a traditional non-neural program executing the same restacking procedure. Moreover, after programming the NVM, the neurocomputational encodings of symbolic block stacking knowledge can be fine-tuned to further improve performance, by applying reinforcement learning to the underlying neural architecture.

show abstract

Section: Introductionmentioning

confidence: 99%