{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:MCEQAYP2I5VGJGSCCVPGLLRXJD","short_pith_number":"pith:MCEQAYP2","schema_version":"1.0","canonical_sha256":"60890061fa476a649a42155e65ae3748e558fda692d867aaf4c4efa2c189a7e6","source":{"kind":"arxiv","id":"2207.01405","version":4},"attestation_state":"computed","paper":{"title":"I-ViT: Integer-only Quantization for Efficient Vision Transformer Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Qingyi Gu, Zhikai Li","submitted_at":"2022-07-04T13:37:38Z","abstract_excerpt":"Vision Transformers (ViTs) have achieved state-of-the-art performance on various computer vision applications. However, these models have considerable storage and computational overheads, making their deployment and efficient inference on edge devices challenging. Quantization is a promising approach to reducing model complexity, and the dyadic arithmetic pipeline can allow the quantized models to perform efficient integer-only inference. Unfortunately, dyadic arithmetic is based on the homogeneity condition in convolutional neural networks, which is not applicable to the non-linear components"},"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":"2207.01405","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2022-07-04T13:37:38Z","cross_cats_sorted":[],"title_canon_sha256":"198ea808800a790ef06cc107bc12e29a3124cc5accb38da2c175777c58f728b5","abstract_canon_sha256":"5381e1a1527d6efe257e67b14641e457e7e4ec4aa94afaf5f0f3b7c7ec8cd0d6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:37:49.856436Z","signature_b64":"xxSoH2zerMQ7fyBsXRj3ZRdBvBQrA1+IUdiCH+p3qKggczZeEKXdYEfzIOhRb0xf9i+nka91pUnDledUxot/Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"60890061fa476a649a42155e65ae3748e558fda692d867aaf4c4efa2c189a7e6","last_reissued_at":"2026-07-05T06:37:49.855902Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:37:49.855902Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"I-ViT: Integer-only Quantization for Efficient Vision Transformer Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Qingyi Gu, Zhikai Li","submitted_at":"2022-07-04T13:37:38Z","abstract_excerpt":"Vision Transformers (ViTs) have achieved state-of-the-art performance on various computer vision applications. However, these models have considerable storage and computational overheads, making their deployment and efficient inference on edge devices challenging. Quantization is a promising approach to reducing model complexity, and the dyadic arithmetic pipeline can allow the quantized models to perform efficient integer-only inference. Unfortunately, dyadic arithmetic is based on the homogeneity condition in convolutional neural networks, which is not applicable to the non-linear components"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2207.01405","kind":"arxiv","version":4},"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/2207.01405/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":"2207.01405","created_at":"2026-07-05T06:37:49.855983+00:00"},{"alias_kind":"arxiv_version","alias_value":"2207.01405v4","created_at":"2026-07-05T06:37:49.855983+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2207.01405","created_at":"2026-07-05T06:37:49.855983+00:00"},{"alias_kind":"pith_short_12","alias_value":"MCEQAYP2I5VG","created_at":"2026-07-05T06:37:49.855983+00:00"},{"alias_kind":"pith_short_16","alias_value":"MCEQAYP2I5VGJGSC","created_at":"2026-07-05T06:37:49.855983+00:00"},{"alias_kind":"pith_short_8","alias_value":"MCEQAYP2","created_at":"2026-07-05T06:37:49.855983+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2304.10891","citing_title":"Transformer-Based Autonomous Driving Models and Deployment-Oriented Compression: A Survey","ref_index":98,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MCEQAYP2I5VGJGSCCVPGLLRXJD","json":"https://pith.science/pith/MCEQAYP2I5VGJGSCCVPGLLRXJD.json","graph_json":"https://pith.science/api/pith-number/MCEQAYP2I5VGJGSCCVPGLLRXJD/graph.json","events_json":"https://pith.science/api/pith-number/MCEQAYP2I5VGJGSCCVPGLLRXJD/events.json","paper":"https://pith.science/paper/MCEQAYP2"},"agent_actions":{"view_html":"https://pith.science/pith/MCEQAYP2I5VGJGSCCVPGLLRXJD","download_json":"https://pith.science/pith/MCEQAYP2I5VGJGSCCVPGLLRXJD.json","view_paper":"https://pith.science/paper/MCEQAYP2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2207.01405&json=true","fetch_graph":"https://pith.science/api/pith-number/MCEQAYP2I5VGJGSCCVPGLLRXJD/graph.json","fetch_events":"https://pith.science/api/pith-number/MCEQAYP2I5VGJGSCCVPGLLRXJD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MCEQAYP2I5VGJGSCCVPGLLRXJD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MCEQAYP2I5VGJGSCCVPGLLRXJD/action/storage_attestation","attest_author":"https://pith.science/pith/MCEQAYP2I5VGJGSCCVPGLLRXJD/action/author_attestation","sign_citation":"https://pith.science/pith/MCEQAYP2I5VGJGSCCVPGLLRXJD/action/citation_signature","submit_replication":"https://pith.science/pith/MCEQAYP2I5VGJGSCCVPGLLRXJD/action/replication_record"}},"created_at":"2026-07-05T06:37:49.855983+00:00","updated_at":"2026-07-05T06:37:49.855983+00:00"}