{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:LF5NTYEWBDJ5CU3T3L4HBCOW5S","short_pith_number":"pith:LF5NTYEW","schema_version":"1.0","canonical_sha256":"597ad9e09608d3d15373daf87089d6ecb439be08c91f99abda60907439f1d33e","source":{"kind":"arxiv","id":"2106.15183","version":3},"attestation_state":"computed","paper":{"title":"Multi-Exit Vision Transformer for Dynamic Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Alexandros Iosifidis, Arian Bakhtiarnia, Qi Zhang","submitted_at":"2021-06-29T09:01:13Z","abstract_excerpt":"Deep neural networks can be converted to multi-exit architectures by inserting early exit branches after some of their intermediate layers. This allows their inference process to become dynamic, which is useful for time critical IoT applications with stringent latency requirements, but with time-variant communication and computation resources. In particular, in edge computing systems and IoT networks where the exact computation time budget is variable and not known beforehand. Vision Transformer is a recently proposed architecture which has since found many applications across various domains "},"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":"2106.15183","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2021-06-29T09:01:13Z","cross_cats_sorted":[],"title_canon_sha256":"75c063761181850574edaaf7e17f22b070e63345650aebebeaded85ee4f41ce7","abstract_canon_sha256":"1367bdf315cf9abdad255439cff952ef4855cdf8a4559a313e99d1b7d7df561f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:24:43.699824Z","signature_b64":"QF0CMygsdI2z2hhi22r2jeP/ac5cGbDnkfmooT22X/oVZmW9RrAKERL1v2E1NBv/9ChcPZ4IxVQQgADozDQWDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"597ad9e09608d3d15373daf87089d6ecb439be08c91f99abda60907439f1d33e","last_reissued_at":"2026-07-05T03:24:43.699301Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:24:43.699301Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Multi-Exit Vision Transformer for Dynamic Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Alexandros Iosifidis, Arian Bakhtiarnia, Qi Zhang","submitted_at":"2021-06-29T09:01:13Z","abstract_excerpt":"Deep neural networks can be converted to multi-exit architectures by inserting early exit branches after some of their intermediate layers. This allows their inference process to become dynamic, which is useful for time critical IoT applications with stringent latency requirements, but with time-variant communication and computation resources. In particular, in edge computing systems and IoT networks where the exact computation time budget is variable and not known beforehand. Vision Transformer is a recently proposed architecture which has since found many applications across various domains "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.15183","kind":"arxiv","version":3},"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/2106.15183/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":"2106.15183","created_at":"2026-07-05T03:24:43.699376+00:00"},{"alias_kind":"arxiv_version","alias_value":"2106.15183v3","created_at":"2026-07-05T03:24:43.699376+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.15183","created_at":"2026-07-05T03:24:43.699376+00:00"},{"alias_kind":"pith_short_12","alias_value":"LF5NTYEWBDJ5","created_at":"2026-07-05T03:24:43.699376+00:00"},{"alias_kind":"pith_short_16","alias_value":"LF5NTYEWBDJ5CU3T","created_at":"2026-07-05T03:24:43.699376+00:00"},{"alias_kind":"pith_short_8","alias_value":"LF5NTYEW","created_at":"2026-07-05T03:24:43.699376+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.07317","citing_title":"Amortized-Precision Quantization for Early-Exit Vision Transformers","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14789","citing_title":"A Comparative Study of CNN Optimization Methods for Edge AI: Exploring the Role of Early Exits","ref_index":52,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LF5NTYEWBDJ5CU3T3L4HBCOW5S","json":"https://pith.science/pith/LF5NTYEWBDJ5CU3T3L4HBCOW5S.json","graph_json":"https://pith.science/api/pith-number/LF5NTYEWBDJ5CU3T3L4HBCOW5S/graph.json","events_json":"https://pith.science/api/pith-number/LF5NTYEWBDJ5CU3T3L4HBCOW5S/events.json","paper":"https://pith.science/paper/LF5NTYEW"},"agent_actions":{"view_html":"https://pith.science/pith/LF5NTYEWBDJ5CU3T3L4HBCOW5S","download_json":"https://pith.science/pith/LF5NTYEWBDJ5CU3T3L4HBCOW5S.json","view_paper":"https://pith.science/paper/LF5NTYEW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2106.15183&json=true","fetch_graph":"https://pith.science/api/pith-number/LF5NTYEWBDJ5CU3T3L4HBCOW5S/graph.json","fetch_events":"https://pith.science/api/pith-number/LF5NTYEWBDJ5CU3T3L4HBCOW5S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LF5NTYEWBDJ5CU3T3L4HBCOW5S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LF5NTYEWBDJ5CU3T3L4HBCOW5S/action/storage_attestation","attest_author":"https://pith.science/pith/LF5NTYEWBDJ5CU3T3L4HBCOW5S/action/author_attestation","sign_citation":"https://pith.science/pith/LF5NTYEWBDJ5CU3T3L4HBCOW5S/action/citation_signature","submit_replication":"https://pith.science/pith/LF5NTYEWBDJ5CU3T3L4HBCOW5S/action/replication_record"}},"created_at":"2026-07-05T03:24:43.699376+00:00","updated_at":"2026-07-05T03:24:43.699376+00:00"}