# Hybrid-NL2SVA: Integrating RAG and Finetuning for LLM-based NL2SVA

arXiv: https://arxiv.org/abs/2506.21569 · HTML: https://mlsanity.com/paper/2506.21569/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Placed in RAG pipeline design and evaluation, one of 15 directions in Retrieval-augmented generation.
- Builds on 16 of its 34 references that are in the map.
- First posted on arXiv 12 Jun 2025.

## Placed in

- Retrieval-augmented generation / [RAG pipeline design and evaluation](https://mlsanity.com/rag/rag-pipeline-design-and-evaluation/)

## Builds on (most cited first, 16 of 16)

- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [LlamaFactory: Unified Efficient Fine-Tuning of 100+ Language Models](https://mlsanity.com/paper/2403.13372/) (2024)
- [Qwen2.5-Coder Technical Report](https://mlsanity.com/paper/2409.12186/) (2024)
- [Benchmarking Retrieval-Augmented Generation for Medicine](https://mlsanity.com/paper/2402.13178/) (2024)
- [nl2spec: Interactively Translating Unstructured Natural Language to Temporal Logics with Large Language Models](https://mlsanity.com/paper/2303.04864/) (2023)
- [Improving Retrieval for RAG based Question Answering Models on Financial Documents](https://mlsanity.com/paper/2404.07221/) (2024)
- [Don't Forget to Connect! Improving RAG with Graph-based Reranking](https://mlsanity.com/paper/2405.18414/) (2024)
- [MAIN-RAG: Multi-Agent Filtering Retrieval-Augmented Generation](https://mlsanity.com/paper/2501.00332/) (2024)
- [AssertLLM: Generating Hardware Verification Assertions from Design Specifications via Multi-LLMs](https://mlsanity.com/paper/2411.14436/) (2024)
- [ChunkRAG: Novel LLM-Chunk Filtering Method for RAG Systems](https://mlsanity.com/paper/2410.19572/) (2024)
- [ChIRAAG: ChatGPT Informed Rapid and Automated Assertion Generation](https://mlsanity.com/paper/2402.00093/) (2024)
- [AssertionBench: A Benchmark to Evaluate Large-Language Models for Assertion Generation](https://mlsanity.com/paper/2406.18627/) (2024)
- [FVEval: Understanding Language Model Capabilities in Formal Verification of Digital Hardware](https://mlsanity.com/paper/2410.23299/) (2024)
- [Toward Conversational Agents with Context and Time Sensitive Long-term Memory](https://mlsanity.com/paper/2406.00057/) (2024)
- [Spec2Assertion: Automatic Pre-RTL Assertion Generation using Large Language Models with Progressive Regularization](https://mlsanity.com/paper/2505.07995/) (2025)
- [Automatic High-quality Verilog Assertion Generation through Subtask-Focused Fine-Tuned LLMs and Iterative Prompting](https://mlsanity.com/paper/2411.15442/) (2024)

## Built on by (most cited first, 6 of 6)

- [QiMeng-CodeV-SVA: Training Specialized LLMs for Hardware Assertion Generation via RTL-Grounded Bidirectional Data Synthesis](https://mlsanity.com/paper/2603.14239/) (2026)
- [GUIDE: GenAI Units In Digital Design Education](https://mlsanity.com/paper/2603.17296/) (2026)
- [FVRuleLearner: Operator-Level Reasoning Tree (Op-Tree)-Based Rules Learning for Formal Verification](https://mlsanity.com/paper/2604.03245/) (2026)
- [LLMs for Secure Hardware Design and Related Problems: Opportunities and Challenges](https://mlsanity.com/paper/2605.10807/) (2026)
- [Hardware Design and Security in the Era of Chiplets and LLMs](https://mlsanity.com/paper/2608.05063/) (2026)
- [A Survey on LLM-Integrated Hardware Design Verification](https://mlsanity.com/paper/2610.10580/) (2026)

## Abstract (arXiv)

SystemVerilog Assertions (SVAs) are critical for verifying the correctness of hardware designs, but manually writing them from natural language property descriptions, i.e., NL2SVA, remains a labor-intensive and error-prone task. Recent advances in large language models (LLMs) offer opportunities to automate this translation. However, existing models still struggle with understanding domain-specific syntax and semantics. To enhance LLM performance in NL2SVA, we propose a customized retrieval-augmented generation (RAG) framework and a synthetic fine-tuning dataset that together improve LLM's performance. To further improve lightweight models over NL2SVA, our fine-tuning dataset provides prompt-guided explanations that teach LLMs the layer-by-layer construction process of concurrent SVAs, enabling supervised fine-tuning that greatly improves syntax and functionality accuracy. To evaluate the performance of LLMs over NL2SVA, we construct the largest evaluation dataset for NL2SVA, comprising 40 Verilog designs and 229 formally verified SVAs with detailed annotations. Experimental results show that our customized RAG framework increases the number of functionality matched SVAs by 58.42% over GPT-4o-mini, while Qwen2.5-Coder-7B-Instruct fine-tuned on our fine-tuning dataset and integrated with HybridRetrieval achieves a 59.05% over the base Qwen model.