Passport: Improving Automated Formal Verification Using Identifiers

Formally verifying system properties is one of the most effective ways improving quality, but its high manual effort requirements often render it prohibitively expensive. Tools that automate formal verification by learning from proof corpora to synthesize proofs have just begun show their promise. These tools are because richness data contain. This comes stylistic conventions followed communities developers, together with powerful logical systems beneath assistants. However, this remains underexploited, work thus far focusing on architecture rather than how make data. article systematically explores effectively exploit aspect data: identifiers. We develop Passport approach, a method for enriching predictive Coq model used an existing proof-synthesis tool three new encoding mechanisms identifiers: category vocabulary indexing, subword sequence modeling, and path elaboration. evaluate our approach’s enrichment effect base tools: ASTactic, Tac, Tok. In head-to-head comparisons, automatically proves 29% more theorems best-performing these tools. Combining enhanced approach 38% combining Finally, together, versions prove 45% combined Overall, findings suggest modeling identifiers can play significant role in synthesis, leading higher-quality software.