PulseAugur
EN
LIVE 08:21:37

LLMs drive neural architecture search with new methods for code and mobile deployment

By PulseAugur Editorial · [2 sources] ·

Two new research papers explore the use of Large Language Models (LLMs) in Neural Architecture Search (NAS). The first paper, 'GraphIR', introduces an intermediate representation to bridge the gap between executable neural network programs and LLM-guided evolution, showing improved search performance on benchmarks like CLRS. The second paper, 'Device-First Feedback', focuses on mobile-native LLM-driven NAS, developing a pipeline that optimizes models for on-device performance rather than just GPU accuracy, demonstrating significant gains on mobile deployment scores for CIFAR-10 but facing challenges with harder tasks like CIFAR-100. AI

IMPACT These papers advance LLM capabilities in automating complex model design, potentially accelerating AI development and deployment.

RANK_REASON Two academic papers published on arXiv detailing new methods for LLM-driven Neural Architecture Search.

Read on arXiv cs.LG →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

LLMs drive neural architecture search with new methods for code and mobile deployment

COVERAGE [2]

  1. arXiv cs.LG TIER_1 English(EN) · Zhen Liu, Wanqi Zhou, Shuanghao Bai, Yuhan Liu, Jinjun Wang, Jingwen Fu ·

    GraphIR: Architecture-Level Search States for LLM-Guided Neural Architecture Evolution

    arXiv:2608.01633v1 Announce Type: new Abstract: Large language models (LLMs) enable neural architecture search (NAS) directly over executable neural network programs. However, code-level flexibility does not provide the architecture state needed for effective mutation: LLMs must …

  2. arXiv cs.CV TIER_1 English(EN) · Saif U Din, Muhammad Ahsan Hussain, Radu Timofte, Dmitry Ignatov ·

    Device-First Feedback: Toward Mobile-Native LLM-Driven Neural Architecture Search

    arXiv:2608.00078v1 Announce Type: new Abstract: Deploying convolutional neural networks generated by large language models (LLMs) on real mobile hardware requires more than GPU validation accuracy: INT8 TensorFlow Lite export, delegate selection, and on-device latency jointly det…

RELATED ENTITIES

RELATED TOPICS