GrepSeek: Training Search Agents for Direct Corpus Interaction

• URL: http://arxiv.org/abs/2605.29307v1
• Date: Thu, 28 May 2026 03:37:33 GMT
• Title: GrepSeek: Training Search Agents for Direct Corpus Interaction
• Abstract summary: GrepSeek is an optimized direct corpus interaction (DCI) search agent that trains a compact search agent to find, filter, and compose evidence from large text corpora.<n>To make DCI practical at scale, we further use a semantics-preserving sharded-parallel execution engine that accelerates shell-based retrieval by up to $7.6times$ while preserving byte-exact equivalence with sequential execution of the shell command.
• Score: 66.69568141699311
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large Language Model (LLM) search agents have shown strong promise for knowledge-intensive language tasks through multiple rounds of reasoning and information retrieval. Most existing systems access information using a retriever that takes a keyword or natural language query and returns a ranked list of documents using an index of pre-computed document representations. In this work, we explore a complementary perspective in which the search agent treats the corpus itself as the search environment and finds evidence by issuing executable shell commands. We introduce GrepSeek, an optimized direct corpus interaction (DCI) search agent that trains a compact search agent to find, filter, and compose evidence from large text corpora. To address the instability of learning behavior directly with reinforcement learning on large corpora, we propose a two-stage training pipeline. First, we construct a cold-start dataset using an answer-aware Tutor and answer-blind Planner to generate verified, causally grounded search trajectories. Second, we refine the initialized policy with Group Relative Policy Optimization (GRPO), allowing the agent to improve its task-oriented search behavior through direct interaction with the corpus. To make DCI practical at scale, we further use a semantics-preserving sharded-parallel execution engine that accelerates shell-based retrieval by up to $7.6\times$ while preserving byte-exact equivalence with sequential execution of the shell command. Experiments across seven open-domain question answering benchmarks show that GrepSeek achieves the strongest overall token-level $F_1$ and Exact Match. Our analysis also highlights the limitations of purely lexical interaction on queries with substantial surface-form variation, suggesting DCI as a practical and competitive method for search agents that can complement existing retrieval paradigms in the real world.

Related papers

• A New Role for Relevance: Guiding Corpus Interaction in Agentic Search [45.65771530564621]
  Relevance is a query-dependent estimate of whether a document or excerpt contains useful evidence.<n>We introduce the Relevance-Aware RipGrep Search Agent (RARG), which turns relevance into an execution prior to interaction.
  arXiv  Detail & Related papers  (2026-07-27T09:56:52Z)
• SearchOS-V1: Towards Robust Open-Domain Information-Seeking Agent Collaboration [77.28003194268977]
  SearchOS is a system-level multi-agent framework that turns fragile, implicit search progress into explicit, persistent, and shared state.<n>Built on SOCM, SearchOS applies a pipeline-parallel scheduling mechanism that overlaps the execution of sub-agents.<n>SearchOS leads all metrics among the evaluated single- and multi-agent baselines.
  arXiv  Detail & Related papers  (2026-07-16T17:51:23Z)
• GRASP: GRanularity-Aware Search Policy for Agentic RAG [35.781196496632035]
  We introduce GRASP, a reinforcement learning framework for training agents to adaptively coordinate complementary retrieval tools.<n> GRASP provides the agent with semantic search, keyword search, and paragraph-reading actions.<n>We train the policy with a reward that jointly accounts for answer accuracy, grounded reading, complementary search, and turn efficiency.
  arXiv  Detail & Related papers  (2026-07-11T19:58:07Z)
• Is Grep All You Need? How Agent Harnesses Reshape Agentic Search [0.7199239000118145]
  This paper reports an empirical study organized into two experiments.<n>It shows that Ch generally yields higher accuracy than vector retrieval in our comparisons in experiment 1.<n>Overall scores still depend strongly on which harness and tool-calling style is used, even when the underlying conversation data are the same.
  arXiv  Detail & Related papers  (2026-05-14T17:58:41Z)
• Superintelligent Retrieval Agent: The Next Frontier of Information Retrieval [25.731213365755234]
  We introduce textitSuperIntelligent Retrieval Agent (SIRA)<n>SIRA can compress multi-round exploratory search into a single corpus-discriminative retrieval action.<n>It can exceed substantially more expensive multi-round search while remaining interpretable, training-free, and efficient.
  arXiv  Detail & Related papers  (2026-05-07T17:54:29Z)
• Beyond Semantic Similarity: Rethinking Retrieval for Agentic Search via Direct Corpus Interaction [127.64173950476702]
  We study direct corpus interaction (DCI), where an agent searches the raw corpus directly with general-purpose terminal tools.<n>This approach requires no offline indexing and adapts naturally to evolving local corpora.<n>Across IR benchmarks and end-to-end agentic search tasks, this simple setup substantially outperforms strong sparse, dense, and reranking baselines.
  arXiv  Detail & Related papers  (2026-05-03T19:13:11Z)
• Bridging the Pose-Semantic Gap: A Cascade Framework for Text-Based Person Anomaly Search [45.34874989015716]
  Text-based person anomaly search retrieves specific behavioral events from surveillance archives using natural-language queries.<n>Recent pose-aware methods face a fundamental Pose-Semantic Gap: semantically different actions can share similar skeletal geometries.<n>We propose the Structure-Semantic Decoupled Cascade (SSDC) framework, which decouples retrieval into two stages.
  arXiv  Detail & Related papers  (2026-04-25T12:53:15Z)
• Learning Refined Document Representations for Dense Retrieval via Deliberate Thinking [58.69615583599489]
  Deliberate Thinking based Retriever (Debater) is a novel approach that enhances document representations by incorporating a step-by-step thinking process.<n>Debater significantly outperforms existing methods across several retrieval benchmarks.
  arXiv  Detail & Related papers  (2025-02-18T15:56:34Z)
• CorpusLM: Towards a Unified Language Model on Corpus for Knowledge-Intensive Tasks [20.390672895839757]
  Retrieval-augmented generation (RAG) has emerged as a popular solution to enhance factual accuracy. Traditional retrieval modules often rely on large document index and disconnect with generative tasks. We propose textbfCorpusLM, a unified language model that integrates generative retrieval, closed-book generation, and RAG.
  arXiv  Detail & Related papers  (2024-02-02T06:44:22Z)
• Dense X Retrieval: What Retrieval Granularity Should We Use? [56.90827473115201]
  Often-overlooked design choice is the retrieval unit in which the corpus is indexed, e.g. document, passage, or sentence. We introduce a novel retrieval unit, proposition, for dense retrieval. Experiments reveal that indexing a corpus by fine-grained units such as propositions significantly outperforms passage-level units in retrieval tasks.
  arXiv  Detail & Related papers  (2023-12-11T18:57:35Z)
• UnifieR: A Unified Retriever for Large-Scale Retrieval [84.61239936314597]
  Large-scale retrieval is to recall relevant documents from a huge collection given a query. Recent retrieval methods based on pre-trained language models (PLM) can be coarsely categorized into either dense-vector or lexicon-based paradigms. We propose a new learning framework, UnifieR which unifies dense-vector and lexicon-based retrieval in one model with a dual-representing capability.
  arXiv  Detail & Related papers  (2022-05-23T11:01:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.