Cognitive ARC-AGI-3 solver: 6 human-like drives, 10 reasoning modes (incl. simulation physique), 20 domaines physiques, micro-NN experts (580 params), multi-agent architecture.
-
Updated
Jul 10, 2026 - Python
Cognitive ARC-AGI-3 solver: 6 human-like drives, 10 reasoning modes (incl. simulation physique), 20 domaines physiques, micro-NN experts (580 params), multi-agent architecture.
ChoiceBench is a lightweight framework for MCQ evaluation-method research on LLMs, with built-in support for answer-order bias analysis and mitigation methods.
Add a description, image, and links to the arc-challenge topic page so that developers can more easily learn about it.
To associate your repository with the arc-challenge topic, visit your repo's landing page and select "manage topics."