{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:G7CTOBTMG766E4YUYVI47X4DNI","short_pith_number":"pith:G7CTOBTM","schema_version":"1.0","canonical_sha256":"37c537066c37fde27314c551cfdf836a11acab48022da107d2ee4e50125d5488","source":{"kind":"arxiv","id":"2412.19325","version":1},"attestation_state":"computed","paper":{"title":"Performance Control in Early Exiting to Deploy Large Models at the Same Cost of Smaller Ones","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Ioannis Mitliagkas, Joao Monteiro, Mehrnaz Mofakhami, Reza Bayat, Valentina Zantedeschi","submitted_at":"2024-12-26T18:54:32Z","abstract_excerpt":"Early Exiting (EE) is a promising technique for speeding up inference by adaptively allocating compute resources to data points based on their difficulty. The approach enables predictions to exit at earlier layers for simpler samples while reserving more computation for challenging ones. In this study, we first present a novel perspective on the EE approach, showing that larger models deployed with EE can achieve higher performance than smaller models while maintaining similar computational costs. As existing EE approaches rely on confidence estimation at each exit point, we further study the "},"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":"2412.19325","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-26T18:54:32Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"cbfb98e14ba32daa2faa3cd8cd7e81fcc6b8b1971c33087737ba5de5f5c67124","abstract_canon_sha256":"7c89646131e160e8f755b0757384b301447ed87d4367be47fe4417061227b0ab"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:54:33.897222Z","signature_b64":"HLCmzeRKdejY0K0hjStKrQE9xj+W/j5uARHEDlwXT7SarJYVatx5XbBRUoajxQNb7/EK2Cx1QrtCKk1nUsKzBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"37c537066c37fde27314c551cfdf836a11acab48022da107d2ee4e50125d5488","last_reissued_at":"2026-07-05T09:54:33.896767Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:54:33.896767Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Performance Control in Early Exiting to Deploy Large Models at the Same Cost of Smaller Ones","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Ioannis Mitliagkas, Joao Monteiro, Mehrnaz Mofakhami, Reza Bayat, Valentina Zantedeschi","submitted_at":"2024-12-26T18:54:32Z","abstract_excerpt":"Early Exiting (EE) is a promising technique for speeding up inference by adaptively allocating compute resources to data points based on their difficulty. The approach enables predictions to exit at earlier layers for simpler samples while reserving more computation for challenging ones. In this study, we first present a novel perspective on the EE approach, showing that larger models deployed with EE can achieve higher performance than smaller models while maintaining similar computational costs. As existing EE approaches rely on confidence estimation at each exit point, we further study the "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.19325","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/2412.19325/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":"2412.19325","created_at":"2026-07-05T09:54:33.896824+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.19325v1","created_at":"2026-07-05T09:54:33.896824+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.19325","created_at":"2026-07-05T09:54:33.896824+00:00"},{"alias_kind":"pith_short_12","alias_value":"G7CTOBTMG766","created_at":"2026-07-05T09:54:33.896824+00:00"},{"alias_kind":"pith_short_16","alias_value":"G7CTOBTMG766E4YU","created_at":"2026-07-05T09:54:33.896824+00:00"},{"alias_kind":"pith_short_8","alias_value":"G7CTOBTM","created_at":"2026-07-05T09:54:33.896824+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/G7CTOBTMG766E4YUYVI47X4DNI","json":"https://pith.science/pith/G7CTOBTMG766E4YUYVI47X4DNI.json","graph_json":"https://pith.science/api/pith-number/G7CTOBTMG766E4YUYVI47X4DNI/graph.json","events_json":"https://pith.science/api/pith-number/G7CTOBTMG766E4YUYVI47X4DNI/events.json","paper":"https://pith.science/paper/G7CTOBTM"},"agent_actions":{"view_html":"https://pith.science/pith/G7CTOBTMG766E4YUYVI47X4DNI","download_json":"https://pith.science/pith/G7CTOBTMG766E4YUYVI47X4DNI.json","view_paper":"https://pith.science/paper/G7CTOBTM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.19325&json=true","fetch_graph":"https://pith.science/api/pith-number/G7CTOBTMG766E4YUYVI47X4DNI/graph.json","fetch_events":"https://pith.science/api/pith-number/G7CTOBTMG766E4YUYVI47X4DNI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/G7CTOBTMG766E4YUYVI47X4DNI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/G7CTOBTMG766E4YUYVI47X4DNI/action/storage_attestation","attest_author":"https://pith.science/pith/G7CTOBTMG766E4YUYVI47X4DNI/action/author_attestation","sign_citation":"https://pith.science/pith/G7CTOBTMG766E4YUYVI47X4DNI/action/citation_signature","submit_replication":"https://pith.science/pith/G7CTOBTMG766E4YUYVI47X4DNI/action/replication_record"}},"created_at":"2026-07-05T09:54:33.896824+00:00","updated_at":"2026-07-05T09:54:33.896824+00:00"}