{"paper":{"title":"LLM-Oriented Information Retrieval: A Denoising-First Perspective","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Denoising to maximize evidence density and verifiability becomes the central task in information retrieval for large language models.","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.IR","authors_text":"Cehao Yang, Fanpu Cao, Hao Liu, Hui Xiong, Liang Sun, Lu Dai, Ziyang Rao","submitted_at":"2026-05-01T08:30:52Z","abstract_excerpt":"Modern information retrieval (IR) is no longer consumed primarily by humans but increasingly by large language models (LLMs) via retrieval-augmented generation (RAG) and agentic search. Unlike human users, LLMs are constrained by limited attention budgets and are uniquely vulnerable to noise; misleading or irrelevant information is no longer just a nuisance, but a direct cause of hallucinations and reasoning failures. In this perspective paper, we argue that denoising-maximizing usable evidence density and verifiability within a context window-is becoming the primary bottleneck across the full"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"denoising—maximizing usable evidence density and verifiability within a context window—is becoming the primary bottleneck across the full information access pipeline","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That LLMs' limited attention budgets and unique vulnerability to noise represent a fundamental paradigm shift in IR that requires a new denoising-first framework, rather than incremental extensions of existing relevance and quality techniques.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Denoising to maximize usable evidence density and verifiability is becoming the primary bottleneck in LLM-oriented information retrieval, conceptualized via a four-stage framework and addressed through a pipeline taxonomy of optimization techniques.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Denoising to maximize evidence density and verifiability becomes the central task in information retrieval for large language models.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"75d40d34c9b699fe154c52af6251feae575b826e381e2a413e6e7a5067a15fb7"},"source":{"id":"2605.00505","kind":"arxiv","version":2},"verdict":{"id":"2e0c95ea-5901-4eb9-9730-df21acbec9c5","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-09T18:50:00.916148Z","strongest_claim":"denoising—maximizing usable evidence density and verifiability within a context window—is becoming the primary bottleneck across the full information access pipeline","one_line_summary":"Denoising to maximize usable evidence density and verifiability is becoming the primary bottleneck in LLM-oriented information retrieval, conceptualized via a four-stage framework and addressed through a pipeline taxonomy of optimization techniques.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That LLMs' limited attention budgets and unique vulnerability to noise represent a fundamental paradigm shift in IR that requires a new denoising-first framework, rather than incremental extensions of existing relevance and quality techniques.","pith_extraction_headline":"Denoising to maximize evidence density and verifiability becomes the central task in information retrieval for large language models."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.00505/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_compliance","ran_at":"2026-05-19T18:06:21.354858Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"6c094512fddda63e4659b38e9d4dd198244ccd9d5891a40a004b49592e58816b"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}