ml-autoresearch

Automate machine learning model optimization by running an autonomous research loop that continuously iterates on model changes. The Skill analyzes model behavior to inform and apply the next adjustment, creating an evidence-based optimization cycle.

Yes, you can incorporate scientific literature into your ML model adjustment process by setting the literature parameter to 'on'. This optionally grounds model changes in relevant research, requiring the 'literature-search' skill for integration.

An analysis-first approach for autonomous ML research campaigns prioritizes analyzing actual model behavior to inform the next change. Instead of random trials, it continuously iterates on model adjustments guided by concrete performance evidence.

Yes, you need Python 3.9 or higher to run autonomous ML research loops. The environment requires this specific version to execute the iterative optimization scripts and process the integrated scientific literature correctly.

Optimize a CIFAR-10 model using an autonomous research loop by specifying your target metric and enabling literature integration. The Skill iterates on model changes by analyzing behavior and optionally grounding adjustments in scientific literature.

A limitation of using literature integration for model optimization is the dependency on external tools. You must have the 'literature-search' skill available, and PDF dependencies like pypdf and pdfplumber are required to parse the scientific documents.