Beyond Language Modeling: An Exploration of Multimodal Pretraining
This research tackles the problem of advancing foundation models beyond language for multimodal applications, which is significant for the AI community working on multimodal models.
The authors explored multimodal pretraining, achieving synergy between visual and language data, and demonstrated that a unified multimodal pretraining approach can lead to world modeling capabilities, with a scaling asymmetry where vision requires significantly more data than language. They achieved this through controlled experiments, yielding four key insights, including the effectiveness of Representation Autoencoder (RAE) and Mixture-of-Experts (MoE) architectures.
The visual world offers a critical axis for advancing foundation models beyond language. Despite growing interest in this direction, the design space for native multimodal models remains opaque. We provide empirical clarity through controlled, from-scratch pretraining experiments, isolating the factors that govern multimodal pretraining without interference from language pretraining. We adopt the Transfusion framework, using next-token prediction for language and diffusion for vision, to train on diverse data including text, video, image-text pairs, and even action-conditioned video. Our experiments yield four key insights: (i) Representation Autoencoder (RAE) provides an optimal unified visual representation by excelling at both visual understanding and generation; (ii) visual and language data are complementary and yield synergy for downstream capabilities; (iii) unified multimodal pretraining leads naturally to world modeling, with capabilities emerging from general training; and (iv) Mixture-of-Experts (MoE) enables efficient and effective multimodal scaling while naturally inducing modality specialization. Through IsoFLOP analysis, we compute scaling laws for both modalities and uncover a scaling asymmetry: vision is significantly more data-hungry than language. We demonstrate that the MoE architecture harmonizes this scaling asymmetry by providing the high model capacity required by language while accommodating the data-intensive nature of vision, paving the way for truly unified multimodal models.