From Pixels to Words -- Towards Native One-Vision Models at Scale

Researchers introduced NEO-ov, a native vision-language foundation model that learns pixel-word and cross-frame relationships end to end.

The model removes modular encoders and fusion components, instead aligning pixels to words directly from the input stream.

NEO-ov shows strong fine-grained visual perception and near-parity with modular systems across benchmarks.

The paper also provides architecture ablations and training guidance for building scalable unified multimodal models.