Megadose AI progress, ranked and analyzed.

CORE: Improving Compositional Reasoning in MLLM Embedding via Reranker Distillation

· HF Daily Papers ·
CORE distills a reranker’s compositional judgments into an embedding model to better separate lookalike image-text matches.

The paper targets retrieval failures where models see the same objects but miss different attribute-object bindings. CORE builds candidate lists across five compositional matching levels and trains with a Rank-KL objective to copy the reranker’s ranking. In the authors’ tests, CORE-RERANKER-8B averages 82.7% across COLA, SUGARCREPE++, and NEGBENCH, while CORE-EMBED-8B leads the evaluated embedding models. The gains also transfer to MCMR without hurting COCO and Flickr30K retrieval. HF Daily Papers' note

score 5

Categories: Research