Beyond Semantic Similarity: Rethinking Retrieval for Agentic Search via Direct Corpus Interaction

Hugging Face Daily Papers Papers
agentic-search direct-corpus-interaction retrieval-systems information-retrieval llm-agents

Summary

The paper introduces Direct Corpus Interaction (DCI), a novel approach allowing AI agents to query raw text directly using standard terminal tools instead of traditional embedding-based retrieval. By bypassing fixed similarity interfaces and offline indexing, DCI significantly outperforms conventional sparse, dense, and reranking baselines across multiple IR and agentic search benchmarks.

Modern retrieval systems, whether lexical or semantic, expose a corpus through a fixed similarity interface that compresses access into a single top-k retrieval step before reasoning. This abstraction is efficient, but for agentic search, it becomes a bottleneck: exact lexical constraints, sparse clue conjunctions, local context checks, and multi-step hypothesis refinement are difficult to implement by calling a conventional off-the-shelf retriever, and evidence filtered out early cannot be recovered by stronger downstream reasoning. Agentic tasks further exacerbate this limitation because they require agents to orchestrate multiple steps, including discovering intermediate entities, combining weak clues, and revising the plan after observing partial evidence. To tackle the limitation, we study direct corpus interaction (DCI), where an agent searches the raw corpus directly with general-purpose terminal tools (e.g., grep, file reads, shell commands, lightweight scripts), without any embedding model, vector index, or retrieval API. This approach requires no offline indexing and adapts naturally to evolving local corpora. Across IR benchmarks and end-to-end agentic search tasks, this simple setup substantially outperforms strong sparse, dense, and reranking baselines on several BRIGHT and BEIR datasets, and attains strong accuracy on BrowseComp-Plus and multi-hop QA without relying on any conventional semantic retriever. Our results indicate that as language agents become stronger, retrieval quality depends not only on reasoning ability but also on the resolution of the interface through which the model interacts with the corpus, with which DCI opens a broader interface-design space for agentic search.
Original Article
View Cached Full Text

Cached at: 05/08/26, 02:27 PM

Paper page - Beyond Semantic Similarity: Rethinking Retrieval for Agentic Search via Direct Corpus Interaction

Source: https://huggingface.co/papers/2605.05242 Authors:

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

Abstract

Direct corpus interaction enables more effective agentic search by allowing agents to query raw text directly, outperforming traditional retrieval methods in complex tasks.

Modernretrieval systems, whether lexical or semantic, expose a corpus through a fixed similarity interface that compresses access into a single top-k retrieval step before reasoning. This abstraction is efficient, but foragentic search, it becomes a bottleneck: exact lexical constraints, sparse clue conjunctions, local context checks, and multi-step hypothesis refinement are difficult to implement by calling a conventional off-the-shelf retriever, and evidence filtered out early cannot be recovered by stronger downstream reasoning. Agentic tasks further exacerbate this limitation because they require agents to orchestrate multiple steps, including discovering intermediate entities, combining weak clues, and revising the plan after observing partial evidence. To tackle the limitation, we studydirect corpus interaction(DCI), where an agent searches the raw corpus directly with general-purposeterminal tools(e.g., grep, file reads, shell commands, lightweight scripts), without any embedding model, vector index, or retrieval API. This approach requires no offline indexing and adapts naturally to evolving local corpora. AcrossIR benchmarksand end-to-endagentic searchtasks, this simple setup substantially outperforms strong sparse, dense, andrerankingbaselines on several BRIGHT andBEIR datasets, and attains strong accuracy onBrowseComp-Plusandmulti-hop QAwithout relying on any conventional semantic retriever. Our results indicate that as language agents become stronger, retrieval quality depends not only on reasoning ability but also on the resolution of the interface through which the model interacts with the corpus, with which DCI opens a broader interface-design space foragentic search.

View arXiv pageView PDFGitHub2Add to collection

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.05242 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.05242 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.05242 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Towards Retrieving Interaction Spaces for Agentic Search

Hugging Face Daily Papers

RISE framework constructs bounded interaction spaces for agentic search by combining BM25 retrieval with preprocessed document indexing, enabling efficient corpus exploration while maintaining high accuracy at scale.