SOURCE-LINKED INTELLIGENCE

CORE: Improving Compositional Reasoning in MLLM Embedding via Reranker Distillation

arXiv · AI, language, vision and robotics · article · Sep 3, 2026 · UTC

MLLM-based embedding models remain limited in compositional retrieval, often failing to distinguish scenes containing the same concepts but different attribute-object bindings. Yet the same backbone can resolve such distinctions when used as a cross-attentive reranker, motivating us to distill its compositional judgments into the embedding model. We propose CORE, which synthesizes candidate lists spanning five compositional matching levels and introduces a Rank-KL objective that trains the embedding model to reproduce the reranker's fine-grained ranking. We further introduce a graded evaluatio

Read original source ↗ Open in workspace

recordType

paper

region

Global