{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:J6MJZNMKDW7WRUD37SH2QQ62JZ","short_pith_number":"pith:J6MJZNMK","schema_version":"1.0","canonical_sha256":"4f989cb58a1dbf68d07bfc8fa843da4e7080f62a2bacd935d3ecb0542c88dc3a","source":{"kind":"arxiv","id":"2404.13082","version":2},"attestation_state":"computed","paper":{"title":"Efficient Contextual LLM Cascades through Budget-Constrained Policy Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Carlee Joe-Wong, Ege Onur Taga, Jiasi Chen, Samet Oymak, Xuechen Zhang, Zijian Huang","submitted_at":"2024-04-17T05:56:49Z","abstract_excerpt":"Recent successes in natural language processing have led to the proliferation of large language models (LLMs) by multiple providers. Each LLM offering has different inference accuracy, monetary cost, and latency, and their accuracy further depends on the exact wording of the question (i.e., the specific prompt). At the same time, users often have a limit on monetary budget and latency to answer all their questions, and they do not know which LLMs to choose for each question to meet their accuracy and long term budget requirements. To navigate this rich design space, we propose TREACLE ($\\under"},"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":"2404.13082","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-17T05:56:49Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"4fcb4c4a14fe10512e235cbede7329f199a037be45d40b941e1479961b8b60c2","abstract_canon_sha256":"43ed618e4e719050cf4f6e8675995361e5100ff220f0796324c2ca3bfb1937c3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:37:52.812643Z","signature_b64":"JDmRs4kjx00/t8UVGkjJHfOgS2fyRNnBPTARZvsh0GHgSAo2WrlUhb8YGLuUENpoEaq71SSmEqPutywez/rWBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4f989cb58a1dbf68d07bfc8fa843da4e7080f62a2bacd935d3ecb0542c88dc3a","last_reissued_at":"2026-07-05T09:37:52.812156Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:37:52.812156Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Efficient Contextual LLM Cascades through Budget-Constrained Policy Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Carlee Joe-Wong, Ege Onur Taga, Jiasi Chen, Samet Oymak, Xuechen Zhang, Zijian Huang","submitted_at":"2024-04-17T05:56:49Z","abstract_excerpt":"Recent successes in natural language processing have led to the proliferation of large language models (LLMs) by multiple providers. Each LLM offering has different inference accuracy, monetary cost, and latency, and their accuracy further depends on the exact wording of the question (i.e., the specific prompt). At the same time, users often have a limit on monetary budget and latency to answer all their questions, and they do not know which LLMs to choose for each question to meet their accuracy and long term budget requirements. To navigate this rich design space, we propose TREACLE ($\\under"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.13082","kind":"arxiv","version":2},"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/2404.13082/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":"2404.13082","created_at":"2026-07-05T09:37:52.812211+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.13082v2","created_at":"2026-07-05T09:37:52.812211+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.13082","created_at":"2026-07-05T09:37:52.812211+00:00"},{"alias_kind":"pith_short_12","alias_value":"J6MJZNMKDW7W","created_at":"2026-07-05T09:37:52.812211+00:00"},{"alias_kind":"pith_short_16","alias_value":"J6MJZNMKDW7WRUD3","created_at":"2026-07-05T09:37:52.812211+00:00"},{"alias_kind":"pith_short_8","alias_value":"J6MJZNMK","created_at":"2026-07-05T09:37:52.812211+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.09367","citing_title":"PICE: A Semantic-Driven Progressive Inference System for LLM Serving in Cloud-Edge Networks","ref_index":10,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/J6MJZNMKDW7WRUD37SH2QQ62JZ","json":"https://pith.science/pith/J6MJZNMKDW7WRUD37SH2QQ62JZ.json","graph_json":"https://pith.science/api/pith-number/J6MJZNMKDW7WRUD37SH2QQ62JZ/graph.json","events_json":"https://pith.science/api/pith-number/J6MJZNMKDW7WRUD37SH2QQ62JZ/events.json","paper":"https://pith.science/paper/J6MJZNMK"},"agent_actions":{"view_html":"https://pith.science/pith/J6MJZNMKDW7WRUD37SH2QQ62JZ","download_json":"https://pith.science/pith/J6MJZNMKDW7WRUD37SH2QQ62JZ.json","view_paper":"https://pith.science/paper/J6MJZNMK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.13082&json=true","fetch_graph":"https://pith.science/api/pith-number/J6MJZNMKDW7WRUD37SH2QQ62JZ/graph.json","fetch_events":"https://pith.science/api/pith-number/J6MJZNMKDW7WRUD37SH2QQ62JZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/J6MJZNMKDW7WRUD37SH2QQ62JZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/J6MJZNMKDW7WRUD37SH2QQ62JZ/action/storage_attestation","attest_author":"https://pith.science/pith/J6MJZNMKDW7WRUD37SH2QQ62JZ/action/author_attestation","sign_citation":"https://pith.science/pith/J6MJZNMKDW7WRUD37SH2QQ62JZ/action/citation_signature","submit_replication":"https://pith.science/pith/J6MJZNMKDW7WRUD37SH2QQ62JZ/action/replication_record"}},"created_at":"2026-07-05T09:37:52.812211+00:00","updated_at":"2026-07-05T09:37:52.812211+00:00"}