Pith. sign in

REVIEW 5 cited by

Can GPT-4 Perform Neural Architecture Search?

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2304.10970 v4 pith:HSL3CLQN submitted 2023-04-21 cs.LG

Mingkai Zheng , Xiu Su , Shan You , Fei Wang , Chen Qian , Chang Xu , Samuel Albanie This is my paper
classification cs.LG
keywords textbfgeniusgpt-4architectureneuralsearchcandidatesgithub
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

We investigate the potential of GPT-4~\cite{gpt4} to perform Neural Architecture Search (NAS) -- the task of designing effective neural architectures. Our proposed approach, \textbf{G}PT-4 \textbf{E}nhanced \textbf{N}eural arch\textbf{I}tect\textbf{U}re \textbf{S}earch (GENIUS), leverages the generative capabilities of GPT-4 as a black-box optimiser to quickly navigate the architecture search space, pinpoint promising candidates, and iteratively refine these candidates to improve performance. We assess GENIUS across several benchmarks, comparing it with existing state-of-the-art NAS techniques to illustrate its effectiveness. Rather than targeting state-of-the-art performance, our objective is to highlight GPT-4's potential to assist research on a challenging technical problem through a simple prompting scheme that requires relatively limited domain expertise\footnote{Code available at \href{https://github.com/mingkai-zheng/GENIUS}{https://github.com/mingkai-zheng/GENIUS}.}. More broadly, we believe our preliminary results point to future research that harnesses general purpose language models for diverse optimisation tasks. We also highlight important limitations to our study, and note implications for AI safety.

Discussion (0). Sign in to comment.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. PhaseNAS: Language-Model Driven Architecture Search with Dynamic Phase Adaptation

    cs.LG 2025-07 reject novelty 5.0 of 10

    PhaseNAS uses dynamic small-to-large LLM switching and a template language to search neural architectures, claiming better accuracy and lower search cost on NAS-Bench-Macro, CIFAR, and COCO.

  2. Beyond Model Base Retrieval: Weaving Knowledge to Master Fine-grained Neural Network Design

    cs.LG 2025-07 conditional novelty 5.0 of 10

    M-DESIGN frames neural network refinement as retrieval over modification-gain graphs, with a Bayesian online update of task similarity and predictive planners for out-of-distribution cases, and reports reaching search...

  3. LLM-Driven AutoML for Cross-Lingual Handwritten OCR: Closed-Loop Neural Architecture Search with GPT-5, GPT-4o, and Claude Sonnet 4

    cs.CV 2026-07 reject novelty 4.0 of 10

    An LLM-driven closed-loop neural architecture search is applied to Arabic, Persian, and English handwriting, claiming mean test accuracies above 93% and 41–44 ms inference.

  4. Scaling Closed-Loop Feature Channel Configuration with LLMs

    cs.LG 2026-07 conditional novelty 4.0 of 10

    Scaling LLM-generated channel-configuration search from sparse to 250 candidates per cycle yields a modest mean-accuracy trend, a frontier improvement from 0.3144 to 0.3676, and measurable channel-allocation regularities.

  5. SHeRL-FL: When Representation Learning Meets Split Learning in Hierarchical Federated Learning

    cs.LG 2025-08 unverdicted novelty 2.0 of 10

    The submitted body is an unrelated survey, not the SHeRL-FL method claimed in the metadata.

Pith tools