{"schemaVersion":1,"target":{"type":"reading","id":"f4cc1e9c4d27034fd4934259e052d2e5b1cfea2ac6b3d5d5dc105b7f61cfd42d","revision":"cdc1f958-1a1f-4439-8e8e-54e4e30ba76f","revisionKind":"publication-id","title":"arXiv cs.AI paper proposes Prism for training multimodal models","url":"https://escapemeridian.com/readings/arxiv-cs-ai-paper-proposes-prism-for-training-multimodal-models-f4cc1e9c"},"source":{"type":"source","id":"https://arxiv.org/abs/2611.06343","revision":"9f0d6f3785ba90860d01e391848c7290bf070c532368010004c223ab79bfeaed","revisionKind":"sha256-source-text","title":"Prism: Bidirectional Self-Teaching for Unified Multimodal Models","url":"https://arxiv.org/abs/2611.06343"},"status":"available","items":[],"next":null,"scope":"Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter."}