UEmbed：统一稀疏与稠密多模态嵌入

摘要

稀疏检索支撑着现代搜索系统，从网络搜索到检索增强生成皆是如此。现有工作已引入学习型稀疏检索（LSR），以突破精确词法匹配的局限，迈向更丰富的语义理解。然而，LSR迄今仍局限于编码器式双向架构，其向多模态场景的扩展也仍高度依赖辅助性的跨模态模块。为解决这些局限，我们提出UEmbed（统一嵌入），一种仅解码器的多模态嵌入模型，可在一次因果前向传播中同时生成稀疏词法表示和稠密表示。UEmbed向输入追加N个可学习的特殊标记，并将词表划分为N个互不相交的子集。每个标记的因果隐状态为其所属子集预测稀疏权重，N个子集拼接后形成完整的稀疏向量。我们在公开数据上训练并发布了2B、4B和9B三种规模的UEmbed。UEmbed-9B在MMEB-v2上达到71.8（稠密）和71.0（稀疏）的成绩，超越了基于公开数据训练的多模态嵌入模型（如RzenEmbed）。在BEIR上，UEmbed同样与强力的稠密和稀疏基线保持竞争力。此外，我们从有效性、效率以及智能体应用三个维度展示了UEmbed的实用价值。总体而言，UEmbed提供了一种新范式：它在单一模型中统一了稠密与稀疏嵌入，同时将稀疏检索进一步扩展到文本与多模态输入的统一处理。

English

Sparse retrieval underpins modern search systems, from web search to retrieval-augmented generation. Existing work has introduced Learned Sparse Retrieval (LSR) to push beyond exact lexical matching toward richer semantics. Yet LSR has so far remained tied to encoder-style bidirectional architectures, and its extension to multimodal settings still relies heavily on auxiliary cross-modal modules. To address these limitations, we introduce UEmbed (Unified Embedding), a decoder-only multimodal embedding model that produces both sparse lexical and dense representations in one causal forward pass. UEmbed appends N learnable special tokens to the input and partitions the vocabulary into N disjoint subsets. Each token's causal hidden state predicts sparse weights over its assigned subset, and the N subsets are concatenated into the full sparse vector. Trained on public data, we release UEmbed at 2B, 4B, and 9B scales. UEmbed-9B reaches 71.8 (dense) and 71.0 (sparse) on MMEB-v2, outperforming multimodal embedding models trained on publicly available data (e.g., RzenEmbed). On BEIR, UEmbed also remains competitive with strong dense and sparse baselines. Furthermore, we demonstrate the practical utility of UEmbed across three dimensions: effectiveness, efficiency, and agentic applications. Overall, UEmbed offers a new paradigm: it unifies dense and sparse embeddings in one model, while further extending sparse retrieval to unify text and multimodal inputs.