给单词加个提示词，能显著提升词向量语义相似度表现。
One Word Is Not Enough: Simple Prompts Improve Word Embeddings
- 在单词前加'meaning: {word}'等提示，简单改进嵌入效果。
- 最高提升0.29相关性，部分模型从0提升至0.73以上。
- 无需训练，通用适配各类嵌入模型，适合词义研究者。
文本嵌入模型主要面向句子级任务（如检索、语义相似度），并以句子级基准评估。其在孤立词上的表现尚不明确。我们发现，在词前简单添加语义提示（如'meaning: {word}'或'Represent the semantic concept: {word}'）可显著提升词相似度相关性。在7个标准基准（SimLex-999、WordSim-353、MEN-3000）上测试包括text-embedding-3-large（OpenAI）、embed-english-v3.0（Cohere）、voyage-3（Voyage AI）、all-mpnet-base-v2和Qwen3-Embedding-8B在内的文本嵌入模型，结果显示提示词使SimLex-999的Spearman相关性最高提升0.29。部分模型在无提示时相关性为0，加提示后提升0.73。最佳结果达SimLex-999：0.692（embed-english-v3.0）、WordSim-353：0.811、MEN-3000：0.855（text-embedding-3-large）。优于经典静态嵌入（如Word2Vec：0.40，LexVec：0.48），建立纯嵌入方法新基准。该零样本方法无需训练，兼容任意文本嵌入模型。
原文摘要 · Abstract (English)
Text embedding models are designed for sentence-level applications like retrieval and semantic similarity, and are primarily evaluated on sentence-level benchmarks. Their behavior on isolated words is less understood. We show that simply prepending semantic prompts to words before embedding substantially improves word similarity correlations. Testing 7 text embedding models, including text-embedding-3-large (OpenAI), embed-english-v3.0 (Cohere), voyage-3(Voyage AI), all-mpnet-base-v2, and Qwen3-Embedding-8B, on 3 standard benchmarks (SimLex-999, WordSim-353, MEN-3000), we find that prompts like "meaning: {word}" or "Represent the semantic concept: {word}" improve Spearman correlations by up to +0.29 on SimLex-999. Some models fail completely on bare words (correlation = 0) but recover with prompts (+0.73 improvement). Our best results achieve correlation = 0.692 on SimLex-999 with embed-english-v3.0 (Cohere), correlation = 0.811 on WordSim-353, and correlation = 0.855 on MEN-3000 with text-embedding-3-large (OpenAI). These results outperform classic static embeddings like Word2Vec (correlation = 0.40) and even the best static method LexVec (correlation = 0.48) on SimLex-999, establishing a new state-of-the-art for pure embedding methods. This zero-shot technique requires no training and works with any text embedding model.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。