STELLAR: Structure-guided LLM Assertion Retrieval and Generation for Formal Verification

2025-11-28

AI Summary

Manually writing SystemVerilog Assertions (SVAs) is inefficient and error-prone, while existing large language model (LLM)-based approaches often overlook the structural patterns inherent in hardware designs. This work proposes a structure-aware SVA generation method that encodes RTL modules into abstract syntax tree (AST)-based structural fingerprints, retrieves structurally similar (RTL, SVA) pairs from a knowledge base, and incorporates them into prompt engineering to guide LLMs during assertion synthesis. By introducing, for the first time, a structure-similarity-guided retrieval mechanism, the approach significantly improves the syntactic correctness, stylistic consistency, and functional accuracy of generated assertions, demonstrating the effectiveness of structure-aware retrieval in industrial-scale formal verification.

Technology Category

Machine Learning: Structured LearningConstraint Satisfaction and Optimization: Satisfiability Modulo TheoriesKnowledge Representation and Reasoning: Automated Reasoning and Theorem Proving

Application Category

Search and Retrieval-Augmented AI: Large language models for searchSemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMsGraph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Formal Verification (FV) relies on high-quality SystemVerilog Assertions (SVAs), but the manual writing process is slow and error-prone. Existing LLM-based approaches either generate assertions from scratch or ignore structural patterns in hardware designs and expert-crafted assertions. This paper presents STELLAR, the first framework that guides LLM-based SVA generation with structural similarity. STELLAR represents RTL blocks as AST structural fingerprints, retrieves structurally relevant (RTL, SVA) pairs from a knowledge base, and integrates them into structure-guided prompts. Experiments show that STELLAR achieves superior syntax correctness, stylistic alignment, and functional correctness, highlighting structure-aware retrieval as a promising direction for industrial FV.

Problem

• Formal Verification
• SystemVerilog Assertions
• Structural Similarity
• LLM-based Generation
• Hardware Design

Innovation

• Structure-guided retrieval
• LLM-based assertion generation
• Formal verification
• AST structural fingerprint
• SystemVerilog Assertions