SOURCE-LINKED INTELLIGENCE

Mode Connectivity Beyond Classifiers: Evidence from Generative and Contrastive Models

arXiv · AI, language, vision and robotics · article · Aug 31, 2026 · UTC

The loss landscape of Deep Neural Networks (DNNs) exhibits highly complex and non-convex properties. Recent studies have revealed the phenomenon of mode connectivity, demonstrating that independently trained network modes can be connected via a continuous low-loss path. However, existing mode connectivity research is predominantly confined to classifier-based models, leaving it an open question whether similar geometric properties exist in modern complex models. In this paper, we extend the boundaries of mode connectivity to generative and contrastive domains (specifically DDPM and NanoCLIP).

Read original source ↗ Open in workspace

recordType

paper

region

Global