{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:VCXKP26DPUM2FG3BMDC67CB2VC","short_pith_number":"pith:VCXKP26D","schema_version":"1.0","canonical_sha256":"a8aea7ebc37d19a29b6160c5ef883aa8879abe58583484611df6e7939f712abb","source":{"kind":"arxiv","id":"2310.17177","version":1},"attestation_state":"computed","paper":{"title":"Bridging The Gaps Between Token Pruning and Full Pre-training via Masked Fine-tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Fengyuan Shi, Limin Wang","submitted_at":"2023-10-26T06:03:18Z","abstract_excerpt":"Despite the success of transformers on various computer vision tasks, they suffer from excessive memory and computational cost. Some works present dynamic vision transformers to accelerate inference by pruning redundant tokens. A key to improving token pruning is using well-trained models as initialization for faster convergence and better performance. However, current base models usually adopt full image training, i.e., using full images as inputs and keeping the whole feature maps through the forward process, which causes inconsistencies with dynamic models that gradually reduce tokens, incl"},"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":"2310.17177","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-10-26T06:03:18Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"160cd287252283a3baa78eddde5fd35075d01f32a40e5f900eeebae02428ca67","abstract_canon_sha256":"8e3315f4e9cd64f5a926fc90c0fe8d96b27d757e546972716cca73dd99f4f164"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:05:20.913608Z","signature_b64":"xbnQxbprSES+7q0k9vLtLr43rqpmbAgOpu4SFnugUakg1mYxkDCuquajnAtZY7sd59K+wa5MLomue+3V17b7Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a8aea7ebc37d19a29b6160c5ef883aa8879abe58583484611df6e7939f712abb","last_reissued_at":"2026-07-05T07:05:20.913123Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:05:20.913123Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Bridging The Gaps Between Token Pruning and Full Pre-training via Masked Fine-tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Fengyuan Shi, Limin Wang","submitted_at":"2023-10-26T06:03:18Z","abstract_excerpt":"Despite the success of transformers on various computer vision tasks, they suffer from excessive memory and computational cost. Some works present dynamic vision transformers to accelerate inference by pruning redundant tokens. A key to improving token pruning is using well-trained models as initialization for faster convergence and better performance. However, current base models usually adopt full image training, i.e., using full images as inputs and keeping the whole feature maps through the forward process, which causes inconsistencies with dynamic models that gradually reduce tokens, incl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.17177","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/2310.17177/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":"2310.17177","created_at":"2026-07-05T07:05:20.913182+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.17177v1","created_at":"2026-07-05T07:05:20.913182+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.17177","created_at":"2026-07-05T07:05:20.913182+00:00"},{"alias_kind":"pith_short_12","alias_value":"VCXKP26DPUM2","created_at":"2026-07-05T07:05:20.913182+00:00"},{"alias_kind":"pith_short_16","alias_value":"VCXKP26DPUM2FG3B","created_at":"2026-07-05T07:05:20.913182+00:00"},{"alias_kind":"pith_short_8","alias_value":"VCXKP26D","created_at":"2026-07-05T07:05:20.913182+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/VCXKP26DPUM2FG3BMDC67CB2VC","json":"https://pith.science/pith/VCXKP26DPUM2FG3BMDC67CB2VC.json","graph_json":"https://pith.science/api/pith-number/VCXKP26DPUM2FG3BMDC67CB2VC/graph.json","events_json":"https://pith.science/api/pith-number/VCXKP26DPUM2FG3BMDC67CB2VC/events.json","paper":"https://pith.science/paper/VCXKP26D"},"agent_actions":{"view_html":"https://pith.science/pith/VCXKP26DPUM2FG3BMDC67CB2VC","download_json":"https://pith.science/pith/VCXKP26DPUM2FG3BMDC67CB2VC.json","view_paper":"https://pith.science/paper/VCXKP26D","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.17177&json=true","fetch_graph":"https://pith.science/api/pith-number/VCXKP26DPUM2FG3BMDC67CB2VC/graph.json","fetch_events":"https://pith.science/api/pith-number/VCXKP26DPUM2FG3BMDC67CB2VC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VCXKP26DPUM2FG3BMDC67CB2VC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VCXKP26DPUM2FG3BMDC67CB2VC/action/storage_attestation","attest_author":"https://pith.science/pith/VCXKP26DPUM2FG3BMDC67CB2VC/action/author_attestation","sign_citation":"https://pith.science/pith/VCXKP26DPUM2FG3BMDC67CB2VC/action/citation_signature","submit_replication":"https://pith.science/pith/VCXKP26DPUM2FG3BMDC67CB2VC/action/replication_record"}},"created_at":"2026-07-05T07:05:20.913182+00:00","updated_at":"2026-07-05T07:05:20.913182+00:00"}