{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UT7S4JRCYA24NDHYTBKCERE4UJ","short_pith_number":"pith:UT7S4JRC","schema_version":"1.0","canonical_sha256":"a4ff2e2622c035c68cf8985422449ca244e9f288967e084b9e48acb8e71ecebc","source":{"kind":"arxiv","id":"2607.25346","version":1},"attestation_state":"computed","paper":{"title":"The Case Against Generation for Retrieval: Discriminative Language Models as Effective Retrievers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Anil Ramakrishna, Ankit Peshin, Carl Hu, Chiyu Zhang, Feng Qi, Frank Shyu, Justin Johnson, Kaushik Rangadurai, Kavosh Asadi, Luke Simon, Mingfu Liang, Oladipo Ositelu, Prachi Agrawal, Sandeep Pandey, Tianyi Chen, Tik On Lui, Wuwei Lan, Xi Liu, Yang Yang, Zhe Xu","submitted_at":"2026-07-28T06:47:53Z","abstract_excerpt":"Large Language Models (LLMs) have emerged as powerful assets for recommender systems. However, deploying them as generative recommenders or zero-shot rankers at web-scale remains bottlenecked by prohibitive computational overhead and grounding challenges. In this paper, we revitalize the classic, highly efficient two-tower retrieval architecture by adapting LLMs as semantic representation backbones rather than generative engines. We introduce an LLM-native two-tower framework engineered for high-throughput, large-scale retrieval. Our architecture introduces several key innovations: a shared LL"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2607.25346","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-07-28T06:47:53Z","cross_cats_sorted":[],"title_canon_sha256":"f31e1215cdddb5a2a9f01c28b993bc6616cdc877b5fbaa2fb9575563df522d5a","abstract_canon_sha256":"b0e7593ec54976283de5b07ef2e42dc8c050ea6808889b38fb2873d4c6ab6e7e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-29T01:25:10.609214Z","signature_b64":"wnZ8cEPeXpc5PkSb1xbR6CoxKdDBB29Gjo600u880hqKcQVHHUsdMTncC6gDpxtvazOQe2iV/wx1Jye801RQDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a4ff2e2622c035c68cf8985422449ca244e9f288967e084b9e48acb8e71ecebc","last_reissued_at":"2026-07-29T01:25:10.608351Z","signature_status":"signed_v1","first_computed_at":"2026-07-29T01:25:10.608351Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Case Against Generation for Retrieval: Discriminative Language Models as Effective Retrievers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Anil Ramakrishna, Ankit Peshin, Carl Hu, Chiyu Zhang, Feng Qi, Frank Shyu, Justin Johnson, Kaushik Rangadurai, Kavosh Asadi, Luke Simon, Mingfu Liang, Oladipo Ositelu, Prachi Agrawal, Sandeep Pandey, Tianyi Chen, Tik On Lui, Wuwei Lan, Xi Liu, Yang Yang, Zhe Xu","submitted_at":"2026-07-28T06:47:53Z","abstract_excerpt":"Large Language Models (LLMs) have emerged as powerful assets for recommender systems. However, deploying them as generative recommenders or zero-shot rankers at web-scale remains bottlenecked by prohibitive computational overhead and grounding challenges. In this paper, we revitalize the classic, highly efficient two-tower retrieval architecture by adapting LLMs as semantic representation backbones rather than generative engines. We introduce an LLM-native two-tower framework engineered for high-throughput, large-scale retrieval. Our architecture introduces several key innovations: a shared LL"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.25346","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2607.25346/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2607.25346","created_at":"2026-07-29T01:25:10.608800+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.25346v1","created_at":"2026-07-29T01:25:10.608800+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.25346","created_at":"2026-07-29T01:25:10.608800+00:00"},{"alias_kind":"pith_short_12","alias_value":"UT7S4JRCYA24","created_at":"2026-07-29T01:25:10.608800+00:00"},{"alias_kind":"pith_short_16","alias_value":"UT7S4JRCYA24NDHY","created_at":"2026-07-29T01:25:10.608800+00:00"},{"alias_kind":"pith_short_8","alias_value":"UT7S4JRC","created_at":"2026-07-29T01:25:10.608800+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UT7S4JRCYA24NDHYTBKCERE4UJ","json":"https://pith.science/pith/UT7S4JRCYA24NDHYTBKCERE4UJ.json","graph_json":"https://pith.science/api/pith-number/UT7S4JRCYA24NDHYTBKCERE4UJ/graph.json","events_json":"https://pith.science/api/pith-number/UT7S4JRCYA24NDHYTBKCERE4UJ/events.json","paper":"https://pith.science/paper/UT7S4JRC"},"agent_actions":{"view_html":"https://pith.science/pith/UT7S4JRCYA24NDHYTBKCERE4UJ","download_json":"https://pith.science/pith/UT7S4JRCYA24NDHYTBKCERE4UJ.json","view_paper":"https://pith.science/paper/UT7S4JRC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.25346&json=true","fetch_graph":"https://pith.science/api/pith-number/UT7S4JRCYA24NDHYTBKCERE4UJ/graph.json","fetch_events":"https://pith.science/api/pith-number/UT7S4JRCYA24NDHYTBKCERE4UJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UT7S4JRCYA24NDHYTBKCERE4UJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UT7S4JRCYA24NDHYTBKCERE4UJ/action/storage_attestation","attest_author":"https://pith.science/pith/UT7S4JRCYA24NDHYTBKCERE4UJ/action/author_attestation","sign_citation":"https://pith.science/pith/UT7S4JRCYA24NDHYTBKCERE4UJ/action/citation_signature","submit_replication":"https://pith.science/pith/UT7S4JRCYA24NDHYTBKCERE4UJ/action/replication_record"}},"created_at":"2026-07-29T01:25:10.608800+00:00","updated_at":"2026-07-29T01:25:10.608800+00:00"}