超越Top-K：用可解释智能体操作取代黑盒检索
Beyond Top-K: Replacing Black-Box Retrieval with Interpretable Agentic Operations
论文提出READ，在780页财务报告上准确率58.8%，比向量检索的15.7%高出一大截，每次检索还能当成审计轨迹回放。
论文提出READ（Reliable Embedding-free Agentic Document-search），用三种确定性操作——归一化词法搜索、结构导航、有界跨度读取——替代传统Top-K嵌入检索。
在780页政府财务报告上，86.8%的内容行是表格行，数值单位依赖上方中位数13行的表头，导致分块经常切断数字与其单位的关联。
READ在51个验证问题上达到58.8%准确率，而密集检索仅15.
7%（p_Holm=2×10^-5）。控制实验显示，使用Top-K工具的同款智能体只能达到27.
5%，说明增益来自接口而非迭代。作者也报告BM25与READ在统计上无差异，因此结果区分的是嵌入型与非嵌入型检索，而非智能体与词法检索。
Beyond Top-K: Replacing Black-Box Retrieval with Interpretable Agentic Operations
Retrieval-augmented generation over long documents is dominated by one design: chunk the text, embed the chunks, and surface the top-k nearest neighbours of the query. We argue that for an important class of documents -- financial statements, audit reports, regulatory returns -- this design is structurally unsound, and we make the argument measurable. On a 780-page government financial report, 86.8% of content lines are table rows, thousands of near-identical figures compete in one embedding space, and a figure inherits its unit from a header a median of 13 lines above it -- so a chunk boundary routinely separates a number from whether it is in lakh or crore, an error of two orders of magnitude. A table-aware chunker built as a steelman fixes the unit problem but leaves 27-30% of numeric chunks with no fiscal-year header at every chunk size we tried. We propose READ (Reliable Embedding-free Agentic Document-search), in which an agent reads the raw document through three deterministic operations -- normalized lexical search, structural navigation, and bounded span reads -- exposed over the Model Context Protocol, so a trajectory is a replayable audit trail, not an opaque similarity score. On 51 verified questions READ answers 58.8% against dense retrieval's 15.7% (p_Holm = 2 x 10^-5) -- or 35.3% tuned, which READ still leads by 23.5 points (p_Holm = 0.017). An agent given the same loop but a top-k tool reaches only 27.5%, locating the gain in the interface rather than in iteration. We also report what the evidence does not support: BM25 is statistically indistinguishable from READ, so our result separates embedding-based from embedding-free retrieval, not agentic from lexical search.
This report is categorized under AITOP's "Papers" desk. Key entities and technical topics under continuous tracking include READ, RAG, agents, MCP and tools.
Source attribution: curated from arXiv cs.AI (@Sagar Tamang, Ayush Vyas, Tabarakul Hazarika).
