AI Signal 313
Sentence Transformers v6.0 adds MultiVectorEncoder for ColBERT-style late interaction retrieval
↗
Sentence Transformers v6.0 introduces MultiVectorEncoder, a fourth model type for ColBERT-style late interaction retrieval, supporting PyLate, Stanford-NLP ColBERT, and colpali-engine checkpoints.
Multi-vector models preserve token-level matching information that single-vector embeddings average away, improving retrieval quality at the cost of a larger index. They also enable visual document retrieval by matching text queries against page images without OCR, a capability now available through the familiar Sentence Transformers API.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
MultiVectorEncoder loads PyLate, Stanford-NLP ColBERT, and colpali-engine checkpoints directly.
MaxSim scoring compares each query token against every document token, preserving token-level interaction.
Multi-vector models are state of the art for visual document retrieval, matching text queries to page images without OCR.
THE CLUSTER