推荐kingstnap · 2025-10
“The Google one about what are the limits of embedding dimensions also is great”
推荐 Google 这篇关于嵌入维度极限的论文,称很棒
证明单向量嵌入的维度限制了能检索出的文档子集
作者指出,向量嵌入的理论局限并不只出现在不现实的查询上,在非常简单的现实查询里也会遇到。他们借助学习理论中的已知结果,说明可由某个查询返回的 top-k 文档子集数目受嵌入维度限制,并用可自由参数化的嵌入在测试集上直接优化来验证这一点;返回所有文档对需要相对较高的维度。据此构建了 LIMIT 数据集,结果即使是最先进的嵌入模型也在这个任务上失败。
HN 上的人提它来说明嵌入维度限制了检索能力,语料规模变大时准确率会崩,也提醒向量检索对某些场景容量不足。
“The Google one about what are the limits of embedding dimensions also is great”
“There was recently this paper: On the Theoretical Limitations of Embedding-Based Retrieval”
“However, this recent article empirically characterizes the semantic capacity of embedding vectors, finding inadequate capacity for some use cases.”