CAV Paper Reproducibility

•
1.1k
•

Strengthens research reproducibility for Computer Aided Verification (CAV) papers by standardizing benchmark provenance, tool version pinning, and checkable proof witnesses.

CAV Paper Reproducibility guides researchers through the rigorous validation workflow required for Computer Aided Verification submissions and camera-ready papers. It maps claims directly to verifiable evidence, pins benchmark suites and tool configurations, validates resource limits, and ensures that soundness claims include independent checkable witnesses like DRAT or Isabelle/Coq proofs. Designed for academic and industrial formal verification researchers, this skill eliminates common reviewer-flagged reproducibility pitfalls and ensures full consistency between paper tables and reproducible evaluation artifacts.

主な機能

01Maps theorems, technique claims, and benchmark figures to verifiable artifact locations
02Reconciles paper tables with logged artifact outputs prior to submission and camera-ready deadlines
03Standardizes provenance pinning for SV-COMP, SMT-COMP, HWMCC, and VNN-COMP benchmark suites
04Audits hardware parameters, per-instance resource limits, and portfolio solver seeds
05Ensures soundness claims ship with checkable proof witnesses and independent checkers
061,074 GitHub stars

ユースケース

01Preparing formal verification benchmark evaluations and artifacts for CAV conference submissions
02Verifying consistency between paper tables, solver logs, and scripts prior to camera-ready deadlines
03Structuring checkable proof witnesses and standalone checkers for SMT, SAT, and theorem proving claims