English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In-context learning (ICL) involves reasoning from given contextual examples.As more modalities comes, this procedure is becoming more challenging as theinterleaved input modalities convolutes the understanding process. This isexemplified by the observation that multimodal models often struggle toeffectively extrapolate from contextual examples to perform ICL. To addressthese challenges, we introduce MultiModal In-conteXt Tuning (M$^2$IXT), alightweight module to enhance the ICL capabilities of multimodal unifiedmodels. The proposed M$^2$IXT module perceives an expandable context window toincorporate various labeled examples of multiple modalities (e.g., text, image,and coordinates). It can be prepended to various multimodal unified models(e.g., OFA, Unival, LLaVA) of different architectures and trained via amixed-tasks strategy to enable rapid few-shot adaption on multiple tasks anddatasets. When tuned on as little as 50K multimodal data, M$^2$IXT can boostthe few-shot ICL performance significantly (e.g., 18\% relative increase forOFA), and obtained state-of-the-art results across an array of tasks includingvisual question answering, image captioning, visual grounding, and visualentailment, while being considerably small in terms of model parameters (e.g.,$\sim$$20\times$ smaller than Flamingo or MMICL), highlighting the flexibilityand effectiveness of M$^2$IXT as a multimodal in-context learner.

Lightweight In-Context Tuning for Multimodal Unified Models