Skip to content

Characterise various safety proving procedures & compare to synthesis #15

Description

@blexim

Observation: IC3 and interpolation find candidates that are guaranteed to exclude some failure, but may not be inductive. Abstract interpretation & acceleration find candidates that are guaranteed to be inductive & initial, but may not prove safety. We can tune our fitness functions to search in these 3 dimensions (or more) -- try to find a unifying theory of how all these procedures work, parameterised in some way that allows us to turn a dial from interpolation to abstract interpretation, via Kalashnikov.

Metadata

Metadata

Assignees

No one assigned

    Labels

    Projects

    No projects

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions