English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

M3-Embedding: Multi-Lingual, Multi-Functional, Multi-Granularity Text Embeddings via Self-Knowledge Distillation

Forum topic · 小凯 · 2026-07-05

Summary

M3-Embedding is an embedding model introduced in arXiv:2402.03216 by researchers including Jianlv Chen, Shitao Xiao, and Zheng Liu. It is distinguished by three capabilities: multi-linguality, supporting semantic retrieval across more than 100 working languages; multi-functionality, simultaneously performing dense retrieval, multi-vector retrieval, and sparse retrieval; and multi-granularity, processing inputs from short sentences up to long documents of 8,192 tokens. Training relies on a novel self-knowledge distillation approach in which relevance scores from the different retrieval functionalities are integrated as a teacher signal to improve training quality. An optimized batching strategy enables large batch sizes and high training throughput, improving embedding discriminativeness. Experiments show state-of-the-art results on multilingual, cross-lingual, and long-document retrieval benchmarks. This makes M3-Embedding a strong candidate for RAG pipelines, cross-lingual enterprise search, and long-context retrieval systems.

M3-Embedding: Multi-Lingual, Multi-Functional, Multi-Granularity Text Embeddings via Self-Knowledge Distillation

Overview

  • Paper: M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation
  • Authors: Jianlv Chen, Shitao Xiao, Peitian Zhang, Kun Luo, Defu Lian, Zheng Liu
  • Published: 2024-02-05
  • Source: https://arxiv.org/abs/2402.03216
  • Category: Embedding models
  • Key Capabilities

    M3-Embedding is defined by three "multi" properties:

    1. Multi-Linguality: Uniform support for semantic retrieval across 100+ working languages. 2. Multi-Functionality: Simultaneously supports the three common retrieval modes — dense retrieval, multi-vector retrieval, and sparse retrieval — within a single model. 3. Multi-Granularity: Handles inputs ranging from short sentences to long documents of up to 8,192 tokens.

    Technical Contributions

  • Self-knowledge distillation: Relevance scores produced by the different retrieval functionalities are integrated as a teacher signal, improving training quality beyond any single retrieval mode.
  • Optimized batching strategy: Enables large batch sizes and high training throughput, which improves the discriminativeness of the learned embeddings.
  • Results

    The model achieves new state-of-the-art results on multilingual, cross-lingual, and long-document retrieval benchmarks.

    Original Abstract (verbatim)

    > In this paper, we introduce a new embedding model called M3-Embedding, which is distinguished for its versatility in *Multi-Linguality*, *Multi-Functionality*, and *Multi-Granularity*. It provides a uniform support for the semantic retrieval of more than 100 working languages. It can simultaneously accomplish the three common retrieval functionalities: dense retrieval, multi-vector retrieval, and sparse retrieval. Besides, it is also capable of processing inputs of different granularities, spanning from short sentences to long documents of up to 8,192 tokens. The effective training of M3-Embedding presents a series of technical contributions. Notably, we propose a novel self-knowledge distillation approach, where the relevance scores from different retrieval functionalities can be integrated as the teacher signal to enhance the training quality. We also optimize the batching strategy, which enables a large batch size and high training throughput to improve the discriminativeness of embeddings. M3-Embedding exhibits a superior performance in our experiment, leading to new state-of-the-art results on multilingual, cross-lingual, and long-document retrieval benchmarks.

    Why It Matters for Search & RAG Systems

  • A single model covering dense + sparse + multi-vector retrieval simplifies hybrid search stacks and removes the need to maintain separate retrievers.
  • 8,192-token context support makes it suitable for long-document RAG indexing.
  • Broad language coverage reduces the engineering cost of supporting cross-lingual enterprise search.
  • Self-knowledge distillation is a practical recipe: combine the strengths of multiple retrieval modes into one embedding space.
  • Related Entries

  • The Scandinavian Embedding Benchmarks
  • Arctic-Embed 2.0: Multilingual Retrieval Without Compromise
  • BGE-en-ICL / BGE-ICL: Making Text Embedders Few-Shot Learners
  • Beyond Benchmarks: Evaluating Embedding Model Similarity for Retrieval
  • References

  • Original paper: arXiv:2402.03216

Tags

#embeddings#text-embedding#information-retrieval#multilingual#self-knowledge-distillation#dense-retrieval#sparse-retrieval#rag

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178208626