Apply formal equivalent-mutant filtering to mutation audits

Apply formal equivalence checking in the style of MCY to filter behaviorally equivalent mutants from GateTruth and external RTL-benchmark mutant sets before computing simulation-based mutation scores.

Background

The paper's mutation engine currently reports simulation kill rates without detecting or excluding equivalent mutants. Equivalent mutants can never be killed by a testbench and therefore depress measured mutation scores.

The authors contrast their approach with MCY, which uses formal equivalence checking to discard provably identical mutants, and explicitly identify adopting that approach for GateTruth or audited benchmarks as future work.