# Insights from Autoresearch for Solar Panel Segmentation

arXiv: https://arxiv.org/abs/2610.10491 · HTML: https://mlsanity.com/paper/2610.10491/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 4 of its 15 references that are in the map.
- First posted on arXiv 7 Oct 2026.

## Builds on (most cited first, 4 of 4)

- [Rethinking Atrous Convolution for Semantic Image Segmentation](https://mlsanity.com/paper/1706.05587/) (2017)
- [Decoupled Weight Decay Regularization](https://mlsanity.com/paper/1711.05101/) (2017)
- [The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search](https://mlsanity.com/paper/2504.08066/) (2025)
- [Auto-nnU-Net: Towards Automated Medical Image Segmentation](https://mlsanity.com/paper/2505.16561/) (2025)

## Built on by (most cited first, 1 of 1)

- [The Last Remote Sensing Model Built by Humans? Casting Any Earth Observation Task as Recursive Self-Improvement](https://mlsanity.com/paper/2610.18845/) (2026)

## Abstract (arXiv)

This paper investigates AutoResearch, a protocol in which a coding language model edits a training program under a one-hour GPU budget and retains a change only if validation IoU improves. The protocol is applied to photovoltaic panel segmentation on a frozen real-image split, with DeepLabV3--ResNet-50 held fixed. Three campaigns of 24 experiments, using Gemma~4 12B, Qwen3-8B all improve their one-hour baselines, but retained modifications do not transfer across hardware. The Qwen3-8B configuration, trained on real images only, reaches a test IoU of 0.836 versus 0.833 for the reference GAN-augmented schedule. Research repository https://github.com/VU-AIML/automl4eo-autoresearch-segmentation.