Towards Code-Oriented LM Embeddings for Surrogate-Assisted Neural Architecture Search

• URL: http://arxiv.org/abs/2605.15649v1
• Date: Fri, 15 May 2026 06:07:25 GMT
• Title: Towards Code-Oriented LM Embeddings for Surrogate-Assisted Neural Architecture Search
• Abstract summary: We propose a low-cost embedding strategy that leverages the inductive bias of Language Models (LMs) to eliminate these overheads.<n>Experiments on the NAS-Bench-201 and einspace search spaces reveal that raw code inputs yield higher predictive performance than other text-based encodings.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Developing effective surrogates (performance predictors) for Neural Architecture Search (NAS) typically requires expensive fine-tuning or the engineering of complex representations. We propose a low-cost embedding strategy that leverages the inductive bias of Language Models (LMs) to eliminate these overheads. By representing architectures as PyTorch class definition text, we demonstrate that off-the-shelf LMs act as competitive feature extractors without NAS-specialized fine-tuning. The final predictor is constructed by passing the extracted Code-Oriented LM Embeddings (COLE) through a lightweight regression head. We also investigate strategies to improve embedding quality and utilization. Our experiments on the NAS-Bench-201 and einspace search spaces reveal that raw code inputs yield higher predictive performance than other text-based encodings (e.g., ONNX-to-text encodings) when using frozen LMs. We also observe COLE drives superior surrogate-assisted search using the BANANAS algorithm in NAS-Bench-201. When optimizing for CIFAR-100 performance, replacing structural path encodings with COLE for architecture representation allows for a 34% decrease in the evaluation budget required to reach within 1% of the fittest architecture in the search space (by test accuracy). As any neural architecture can be represented as code, these findings establish COLE as a versatile and efficient foundation for advancing NAS.

Related papers

• ONNX-Net: Towards Universal Representations and Instant Performance Prediction for Neural Architectures [60.14199724905456]
  ONNX-Bench is a benchmark consisting of a collection of neural networks in a unified format based on ONNX files.<n> ONNX-Net represents any neural architecture using natural language descriptions acting as an input to a performance predictor.<n>Experiments show strong zero-shot performance across disparate search spaces using only a small amount of pretraining samples.
  arXiv  Detail & Related papers  (2025-10-06T15:43:36Z)
• LM-Searcher: Cross-domain Neural Architecture Search with LLMs via Unified Numerical Encoding [55.5535016040221]
  LM-Searcher is a novel framework for cross-domain neural architecture optimization.<n>Central to our approach is NCode, a universal numerical string representation for neural architectures.<n>Our dataset, encompassing a wide range of architecture-performance pairs, encourages robust and transferable learning.
  arXiv  Detail & Related papers  (2025-09-06T09:26:39Z)
• Arch-LLM: Taming LLMs for Neural Architecture Generation via Unsupervised Discrete Representation Learning [2.981775461282335]
  A common approach involves the use of Variational Autoencoders (VAEs) to map discrete architectures onto a continuous representation space.<n>We introduce a Vector Quantized Variational Autoencoder (VQ-VAE) to learn a discrete latent space more naturally aligned with the discrete neural architectures.<n>Compared to VAE-based methods, our approach improves the generation of valid and unique architectures by over 80% on NASBench-101 and over 8% on NASBench-201.
  arXiv  Detail & Related papers  (2025-03-28T00:56:56Z)
• Encodings for Prediction-based Neural Architecture Search [8.564763702766776]
  Predictor-based methods have substantially enhanced Neural Architecture Search (NAS) optimization. Traditional encodings used an adjacency matrix describing the graph structure of a neural network. We categorize and investigate neural encodings from three main types: structural, learned, and score-based. We present our predictor textbfFLAN: textbfFlow textbfAttention for textbfNAS.
  arXiv  Detail & Related papers  (2024-03-04T21:05:52Z)
• DNA Family: Boosting Weight-Sharing NAS with Block-Wise Supervisions [121.05720140641189]
  We develop a family of models with the distilling neural architecture (DNA) techniques. Our proposed DNA models can rate all architecture candidates, as opposed to previous works that can only access a sub- search space using algorithms. Our models achieve state-of-the-art top-1 accuracy of 78.9% and 83.6% on ImageNet for a mobile convolutional network and a small vision transformer, respectively.
  arXiv  Detail & Related papers  (2024-03-02T22:16:47Z)
• Efficacy of Neural Prediction-Based Zero-Shot NAS [0.04096453902709291]
  We propose a novel approach for zero-shot Neural Architecture Search (NAS) using deep learning. Our method employs Fourier sum of sines encoding for convolutional kernels, enabling the construction of a computational feed-forward graph with a structure similar to the architecture under evaluation. Experimental results show that our approach surpasses previous methods using graph convolutional networks in terms of correlation on the NAS-Bench-201 dataset and exhibits a higher convergence rate.
  arXiv  Detail & Related papers  (2023-08-31T14:54:06Z)
• A General-Purpose Transferable Predictor for Neural Architecture Search [22.883809911265445]
  We propose a general-purpose neural predictor for Neural Architecture Search (NAS) that can transfer across search spaces. Experimental results on NAS-Bench-101, 201 and 301 demonstrate the efficacy of our scheme.
  arXiv  Detail & Related papers  (2023-02-21T17:28:05Z)
• Generalization Properties of NAS under Activation and Skip Connection Search [66.8386847112332]
  We study the generalization properties of Neural Architecture Search (NAS) under a unifying framework. We derive the lower (and upper) bounds of the minimum eigenvalue of the Neural Tangent Kernel (NTK) under the (in)finite-width regime. We show how the derived results can guide NAS to select the top-performing architectures, even in the case without training.
  arXiv  Detail & Related papers  (2022-09-15T12:11:41Z)
• BaLeNAS: Differentiable Architecture Search via the Bayesian Learning Rule [95.56873042777316]
  Differentiable Architecture Search (DARTS) has received massive attention in recent years, mainly because it significantly reduces the computational cost. This paper formulates the neural architecture search as a distribution learning problem through relaxing the architecture weights into Gaussian distributions. We demonstrate how the differentiable NAS benefits from Bayesian principles, enhancing exploration and improving stability.
  arXiv  Detail & Related papers  (2021-11-25T18:13:42Z)
• Weak NAS Predictors Are All You Need [91.11570424233709]
  Recent predictor-based NAS approaches attempt to solve the problem with two key steps: sampling some architecture-performance pairs and fitting a proxy accuracy predictor. We shift the paradigm from finding a complicated predictor that covers the whole architecture space to a set of weaker predictors that progressively move towards the high-performance sub-space. Our method costs fewer samples to find the top-performance architectures on NAS-Bench-101 and NAS-Bench-201, and it achieves the state-of-the-art ImageNet performance on the NASNet search space.
  arXiv  Detail & Related papers  (2021-02-21T01:58:43Z)
• DrNAS: Dirichlet Neural Architecture Search [88.56953713817545]
  We treat the continuously relaxed architecture mixing weight as random variables, modeled by Dirichlet distribution. With recently developed pathwise derivatives, the Dirichlet parameters can be easily optimized with gradient-based generalization. To alleviate the large memory consumption of differentiable NAS, we propose a simple yet effective progressive learning scheme.
  arXiv  Detail & Related papers  (2020-06-18T08:23:02Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.