ML For Hardware Design Interpretability: Challenges and Opportunities
Not on the map. Why?
Abstract (arXiv)
The increasing size and complexity of machine learning (ML) models have driven the growing need for custom hardware accelerators capable of efficiently supporting ML workloads. However, the design of such accelerators remains a time-consuming process, heavily relying on engineers to manually ensure design interpretability through clear documentation and effective communication. Recent advances in large language models (LLMs) offer a promising opportunity to automate these design interpretability tasks, particularly the generation of natural language descriptions for register-transfer level (RTL) code, what we refer to as "RTL-to-NL tasks." In this paper, we examine how design interpretability, particularly in RTL-to-NL tasks, influences the efficiency of the hardware design process. We review existing work adapting LLMs for these tasks, highlight key challenges that remain unaddressed, including those related to data, computation, and model development, and identify opportunities to address them. By doing so, we aim to guide future research in leveraging ML to automate RTL-to-NL tasks and improve hardware design interpretability, thereby accelerating the hardware design process and meeting the increasing demand for custom hardware accelerators in machine learning and beyond.
Key facts. No citing papers in the map yet (data through 13 Oct 2026). Builds on 53 of its 88 references that are in the map. First posted on arXiv 11 Apr 2025.
Builds on
53 of 88 references.
References · 53
- 2020Language Models are Few-Shot Learnerscited by 32,320
- 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,247
- 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,412
- 2020COMET: A Neural Framework for MT Evaluationcited by 1,027
- 2021CodeT5: Identifier-aware Unified Pre-trained Encoder-Decoder Models for Code Understanding and Generationcited by 905
- 2017Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networkscited by 742
- 2020GraphCodeBERT: Pre-training Code Representations with Data Flowcited by 609
- 2023Gated Linear Attention Transformers with Hardware-Efficient Trainingcited by 500
- 2023Benchmarking Large Language Models for News Summarizationcited by 444
- 2024QuIP#: Even Better LLM Quantization with Hadamard Incoherence and Lattice Codebookscited by 367
- 2020Calibration of Pre-trained Transformerscited by 308
- 2024The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bitscited by 305
38 more
- 2024Long-context LLMs Struggle with Long In-context Learningcited by 286
- 2023Using an LLM to Help With Code Understandingcited by 227
- 2023BitNet: Scaling 1-bit Transformers for Large Language Modelscited by 186
- 2023Navigate through Enigmatic Labyrinth A Survey of Chain of Thought Reasoning: Advances, Frontiers and Futurecited by 184
- 2023RTLLM: An Open-Source Benchmark for Design RTL Generation with Large Language Modelcited by 148
- 2020Training with Quantization Noise for Extreme Model Compressioncited by 145
- 2024Byte Latent Transformer: Patches Scale Better Than Tokenscited by 141
- 2023CodePlan: Repository-level Coding using LLMs and Planningcited by 139
- 2020A Transformer-based Approach for Source Code Summarizationcited by 132
- 2023Machine Learning for Synthetic Data Generation: A Reviewcited by 118
- 2022Benchmarking Large Language Models for Automated Verilog RTL Code Generationcited by 115
- 2024Summary of a Haystack: A Challenge to Long-Context LLMs and RAG Systemscited by 87
- 2023RTLCoder: Outperforming GPT-3.5 in Design RTL Generation with Our Open-Source Dataset and Lightweight Solutioncited by 71
- 2024RepoAgent: An LLM-Powered Open-Source Framework for Repository-level Code Documentation Generationcited by 69
- 2024Why Does the Effective Context Length of LLMs Fall Short?cited by 61
- 2024MG-Verilog: Multi-grained Dataset Towards Enhanced LLM-assisted Verilog Generationcited by 46
- 2024Make Every Move Count: LLM-based High-Quality RTL Code Generation Using MCTScited by 43
- 2022Two-stage LLM Fine-tuning with Less Specialization and More Generalizationcited by 41
- 2023Chainpoll: A high efficacy method for LLM hallucination detectioncited by 39
- 2024Towards LLM-Powered Verilog RTL Assistant: Self-Verification and Self-Correctioncited by 39
- 2020Survey of Machine Learning Acceleratorscited by 29
- 2023A Comparative Analysis of Large Language Models for Code Documentation Generationcited by 24
- 2024Unsupervised Evaluation of Code LLMs with Round-Trip Correctnesscited by 23
- 2024Multitask-based Evaluation of Open-Source LLM on Software Vulnerabilitycited by 23
- 2024What Do Learning Dynamics Reveal About Generalization in LLM Reasoning?cited by 23
- 2025DeepRTL: Bridging Verilog Understanding and Generation with a Unified Representation Modelcited by 23
- 2024NLPerturbator: Studying the Robustness of Code LLMs to Natural Language Variationscited by 20
- 2024CreativEval: Evaluating Creativity of LLM-Based Hardware Code Generationcited by 19
- 2024BitNet a4.8: 4-bit Activations for 1-bit LLMscited by 16
- 2023The potential of LLMs for coding with low-resource and domain-specific programming languagescited by 11
- 2024Evaluating the Generalization Ability of Quantized LLMs: Benchmark, Analysis, and Toolboxcited by 11
- 2024Natural language is not enough: Benchmarking multi-modal generative AI for Verilog generationcited by 11
- 2024SpecLLM: Exploring Generation and Review of VLSI Design Specification with Large Language Modelcited by 9
- 2024SA-DS: A Dataset for Large Language Model-Driven AI Accelerator Design Generationcited by 9
- 2024New Solutions on LLM Acceleration, Optimization, and Applicationcited by 7
- 2023A Survey of Source Code Search: A 3-Dimensional Perspectivecited by 6
- 2020Optimising Design Verification Using Machine Learning: An Open Source Solutioncited by 4
- 2024Zero-Shot RTL Code Generation with Attention Sink Augmented Large Language Modelscited by 4
50 most cited of 53 All 53 (JSON)
Built on by
Not cited yet.