Tunable Neural Encoding of a Symbolic Robotic Manipulation Algorithm

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

Research output: Contribution to journalArticlepeer-review

2 Scopus citations


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.

Original languageEnglish (US)
Article number744031
JournalFrontiers in Neurorobotics
StatePublished - Dec 14 2021
Externally publishedYes


  • explainable AI
  • neurosymbolic architectures
  • policy optimization
  • reinforcement learning
  • robotic manipulation

ASJC Scopus subject areas

  • Biomedical Engineering
  • Artificial Intelligence


Dive into the research topics of 'Tunable Neural Encoding of a Symbolic Robotic Manipulation Algorithm'. Together they form a unique fingerprint.

Cite this