{"paper":{"title":"OBLIQ-Bench: Exposing Overlooked Bottlenecks in Modern Retrievers with Latent and Implicit Queries","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Modern retrievers fail to surface most documents matching latent patterns even when reasoning LLMs can verify them once found.","cross_cats":["cs.AI"],"primary_cat":"cs.IR","authors_text":"Devavrat Shah, Diane Tchuindjo, Omar Khattab","submitted_at":"2026-05-07T13:22:49Z","abstract_excerpt":"Retrieval benchmarks are increasingly saturating, but we argue that efficient search is far from a solved problem. We identify a class of queries we call oblique, which seek documents that instantiate a latent pattern, like finding all tweets that express an implicit stance, chat logs that demonstrate a particular failure mode, or transcripts that match an abstract scenario. We study three mechanisms through which obliqueness may arise and introduce OBLIQ-Bench, a suite of five oblique search problems over real long-tail corpora. OBLIQ-Bench exposes an overlooked asymmetry between retrieval an"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"reasoning LLMs reliably recognize latent relevance whenever relevant documents are surfaced, but even sophisticated retrieval pipelines fail to surface most relevant documents in the first place.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the three mechanisms of obliqueness and the five problems in OBLIQ-Bench are representative of important real-world retrieval challenges and that the observed asymmetry is not an artifact of the specific corpora or models chosen.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"OBLIQ-Bench reveals that modern retrievers fail to surface documents for latent and implicit queries even though LLMs reliably recognize relevance when those documents are provided.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Modern retrievers fail to surface most documents matching latent patterns even when reasoning LLMs can verify them once found.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"e807169c6da03c774604d2d5ff01f24cb5edc3feb7752e1490f50a8402c7faaa"},"source":{"id":"2605.06235","kind":"arxiv","version":2},"verdict":{"id":"171f730d-f2af-408a-a849-93e699d40223","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-08T05:49:10.877169Z","strongest_claim":"reasoning LLMs reliably recognize latent relevance whenever relevant documents are surfaced, but even sophisticated retrieval pipelines fail to surface most relevant documents in the first place.","one_line_summary":"OBLIQ-Bench reveals that modern retrievers fail to surface documents for latent and implicit queries even though LLMs reliably recognize relevance when those documents are provided.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the three mechanisms of obliqueness and the five problems in OBLIQ-Bench are representative of important real-world retrieval challenges and that the observed asymmetry is not an artifact of the specific corpora or models chosen.","pith_extraction_headline":"Modern retrievers fail to surface most documents matching latent patterns even when reasoning LLMs can verify them once found."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.06235/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"claim_evidence","ran_at":"2026-05-20T13:02:04.201471Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-20T08:34:50.401851Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T19:01:19.073001Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T12:53:16.664145Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"c4424c590f20d863e7d0951468fadc3f44256548ff6d0863ad82c3ca3bd2ca06"},"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"}