# Scaling Closed-Loop Feature Channel Configuration with LLMs

arXiv: https://arxiv.org/abs/2607.20516 · HTML: https://mlsanity.com/paper/2607.20516/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 18 of its 22 references that are in the map.
- First posted on arXiv 7 Jul 2026.

## Builds on (most cited first, 18 of 18)

- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [DARTS: Differentiable Architecture Search](https://mlsanity.com/paper/1806.09055/) (2018)
- [Neural Architecture Search with Reinforcement Learning](https://mlsanity.com/paper/1611.01578/) (2016)
- [Learning Efficient Convolutional Networks through Network Slimming](https://mlsanity.com/paper/1708.06519/) (2017)
- [AMC: AutoML for Model Compression and Acceleration on Mobile Devices](https://mlsanity.com/paper/1802.03494/) (2018)
- [MetaPruning: Meta Learning for Automatic Neural Network Channel Pruning](https://mlsanity.com/paper/1903.10258/) (2019)
- [Zero-Cost Proxies for Lightweight NAS](https://mlsanity.com/paper/2101.08134/) (2021)
- [EvoPrompting: Language Models for Code-Level Neural Architecture Search](https://mlsanity.com/paper/2302.14838/) (2023)
- [Evaluating the Search Phase of Neural Architecture Search](https://mlsanity.com/paper/1902.08142/) (2019)
- [LLMatic: Neural Architecture Search via Large Language Models and Quality Diversity Optimization](https://mlsanity.com/paper/2306.01102/) (2023)
- [Can GPT-4 Perform Neural Architecture Search?](https://mlsanity.com/paper/2304.10970/) (2023)
- [EcoNAS: Finding Proxies for Economical Neural Architecture Search](https://mlsanity.com/paper/2001.01233/) (2020)
- [LEMUR Neural Network Dataset: Towards Seamless AutoML](https://mlsanity.com/paper/2504.10552/) (2025)
- [Design Principle Transfer in Neural Architecture Search via Large Language Models](https://mlsanity.com/paper/2408.11330/) (2024)
- [AIRNet: Self-Supervised Affine Registration for 3D Medical Images using Neural Networks](https://mlsanity.com/paper/1810.02583/) (2018)
- [SEKI: Self-Evolution and Knowledge Inspiration based Neural Architecture Search via Large Language Models](https://mlsanity.com/paper/2502.20422/) (2025)
- [CoLLM-NAS: Collaborative Large Language Models for Efficient Knowledge-Guided Neural Architecture Search](https://mlsanity.com/paper/2509.26037/) (2025)
- [Closed-Loop LLM Discovery of Non-Standard Channel Priors in Vision Models](https://mlsanity.com/paper/2601.08517/) (2026)

## Abstract (arXiv)

Promising initial results in closed-loop large-language-model-based channel-configuration search demonstrated that neural-network widths can be optimized directly through executable code generation and accuracy feedback. However, those results were obtained from a relatively sparse set of valid evaluations, leaving open whether the observed optimization behavior transfers to a denser sampling regime and whether additional architectural regularities emerge when more generated networks are evaluated. To test this, the same search setting is scaled to 250 candidate networks per fine-tuning cycle. The analysis covers 2000 generated candidates from 8 complete cycles, yielding 462 verified CIFAR-100 evaluations after task and metadata filtering. Per-cycle mean accuracy exhibits a positive linear trend with slope 9.87e-4 (p=0.043), while the high-performing frontier improves more strongly: the best observed accuracy increases from 0.3144 to 0.3676, and both the top-5 and top-10 cycle-level means exhibit positive trends. The scaled run also reveals improved parameter efficiency. The best model reaches 0.3676 with 11.8M parameters, compared with an early high-performing model at 0.3144 with 166.5M parameters. Beyond accuracy, the larger sample exposes architectural regularities that were difficult to assess from sparse observations. Non-power-of-two channel widths occur in 41.8% of verified candidates, and the strongest models share structured channel-allocation patterns characterized by moderate early widths and expanded middle or later blocks. These findings indicate that the channel-search signal observed in the initial study transfers