SOURCE-LINKED INTELLIGENCE

Multi-modal Knowledge Preserving Adapter for Embedding Backward Compatibility

arXiv · AI, language, vision and robotics · article · Sep 15, 2026 · UTC

Upgrading embedding models typically requires expensive database re-indexing, as new query embeddings are incompatible with existing database embeddings. While Backward Compatible Training (BCT) mitigates this by enforcing compatibility during training, existing approaches often require updating the backbone model. This is impractical because of significant training cost, the risk of performance regression, and limited access to proprietary model weights. We introduce Multi-modal Knowledge Preserving Adapter (MKP-Adapter), the first adapter-only BCT approach for Multi-modal Large Language Mode

Read original source ↗ Open in workspace

recordType

paper

region

Global