ModelSenseNova U1.5 / SenseNova-U1.5-8B-MoT-Preview / SenseNova U1.5 Lite
SenseNova-U1.5
An open-source multimodal image editing model released by SenseTime, built on the Native Unified Paradigm and NEO-unify architecture. It integrates a Mixture-of-Transformers (MoT) mechanism and supports multi-reference image fusion and spatial understanding.
Timeline (last 90 days)
Sep 11
SenseNova-U1.5 adopts the "Native Unified Paradigm" and the NEO-unify architecture, jointly training text understanding, image perception, and generation capabilities within a single unified model framework.
Unverified50%