Given as part of the MPRI 2.8 course “Advanced Techniques of Verification”, this lecture series aims to give a sense of what verification of neural networks means and to explore its theoretical possibilities and limitations.

The lecture covers basic automata-theoretic concepts and logical formalisms for the modeling and verification of concurrent and distributed systems.

The lecture covers several aspects of weighted automata, including recognizable series, probabilistic finite automata, and weighted logic.