Official collection of models and datasets for the ICLR 2026 Oral paper "Through the Lens of Contrast: Self-Improving Visual Reasoning in VLMs"