{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:PEUKDK5BCVX7XN3U33HHYFDR6Z","short_pith_number":"pith:PEUKDK5B","schema_version":"1.0","canonical_sha256":"7928a1aba1156ffbb774dece7c1471f64d509b4bdd0c878f1c660b335b6f1ca0","source":{"kind":"arxiv","id":"2506.08908","version":3},"attestation_state":"computed","paper":{"title":"SkipVAR: Accelerating Visual Autoregressive Modeling via Adaptive Frequency-Aware Skipping","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jiajun Li, Linfeng Zhang, Qingyan Wei, Songhua Liu, Xinyu Zhang, Yue Ma","submitted_at":"2025-06-10T15:35:29Z","abstract_excerpt":"Recent studies on Visual Autoregressive (VAR) models have highlighted that high-frequency components, or later steps, in the generation process contribute disproportionately to inference latency. However, the underlying computational redundancy involved in these steps has yet to be thoroughly investigated. In this paper, we conduct an in-depth analysis of the VAR inference process and identify two primary sources of inefficiency: step redundancy and unconditional branch redundancy. To address step redundancy, we propose an automatic step-skipping strategy that selectively omits unnecessary gen"},"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":"2506.08908","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-06-10T15:35:29Z","cross_cats_sorted":[],"title_canon_sha256":"d063b5bfdd4773d0d8f93d9d826bf50c00d77e54e83b34f7aba328418f894ac7","abstract_canon_sha256":"fd48d63186d88cfe95b4736cd7a34d3cad603fcc379d5bdaed5623df604b6a67"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:34:52.017178Z","signature_b64":"LAk3wA2RY25N4LajHyxATR2feeX+CmOMK1WtzajaMVAgLqUEOFUTJnxXGuRllGiykF8MT4F6w0jZSgqbZGJMCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7928a1aba1156ffbb774dece7c1471f64d509b4bdd0c878f1c660b335b6f1ca0","last_reissued_at":"2026-07-05T11:34:52.016651Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:34:52.016651Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SkipVAR: Accelerating Visual Autoregressive Modeling via Adaptive Frequency-Aware Skipping","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jiajun Li, Linfeng Zhang, Qingyan Wei, Songhua Liu, Xinyu Zhang, Yue Ma","submitted_at":"2025-06-10T15:35:29Z","abstract_excerpt":"Recent studies on Visual Autoregressive (VAR) models have highlighted that high-frequency components, or later steps, in the generation process contribute disproportionately to inference latency. However, the underlying computational redundancy involved in these steps has yet to be thoroughly investigated. In this paper, we conduct an in-depth analysis of the VAR inference process and identify two primary sources of inefficiency: step redundancy and unconditional branch redundancy. To address step redundancy, we propose an automatic step-skipping strategy that selectively omits unnecessary gen"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.08908","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/2506.08908/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":"2506.08908","created_at":"2026-07-05T11:34:52.016719+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.08908v3","created_at":"2026-07-05T11:34:52.016719+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.08908","created_at":"2026-07-05T11:34:52.016719+00:00"},{"alias_kind":"pith_short_12","alias_value":"PEUKDK5BCVX7","created_at":"2026-07-05T11:34:52.016719+00:00"},{"alias_kind":"pith_short_16","alias_value":"PEUKDK5BCVX7XN3U","created_at":"2026-07-05T11:34:52.016719+00:00"},{"alias_kind":"pith_short_8","alias_value":"PEUKDK5B","created_at":"2026-07-05T11:34:52.016719+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.00310","citing_title":"Where to Refine, When to Stop: Rethinking Redundancy via Latent Discrepancy for Efficient Visual Autoregressive Generation","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01220","citing_title":"Visual Implicit Autoregressive Modeling","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17286","citing_title":"Depth Adaptive Efficient Visual Autoregressive Modeling","ref_index":36,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PEUKDK5BCVX7XN3U33HHYFDR6Z","json":"https://pith.science/pith/PEUKDK5BCVX7XN3U33HHYFDR6Z.json","graph_json":"https://pith.science/api/pith-number/PEUKDK5BCVX7XN3U33HHYFDR6Z/graph.json","events_json":"https://pith.science/api/pith-number/PEUKDK5BCVX7XN3U33HHYFDR6Z/events.json","paper":"https://pith.science/paper/PEUKDK5B"},"agent_actions":{"view_html":"https://pith.science/pith/PEUKDK5BCVX7XN3U33HHYFDR6Z","download_json":"https://pith.science/pith/PEUKDK5BCVX7XN3U33HHYFDR6Z.json","view_paper":"https://pith.science/paper/PEUKDK5B","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.08908&json=true","fetch_graph":"https://pith.science/api/pith-number/PEUKDK5BCVX7XN3U33HHYFDR6Z/graph.json","fetch_events":"https://pith.science/api/pith-number/PEUKDK5BCVX7XN3U33HHYFDR6Z/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PEUKDK5BCVX7XN3U33HHYFDR6Z/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PEUKDK5BCVX7XN3U33HHYFDR6Z/action/storage_attestation","attest_author":"https://pith.science/pith/PEUKDK5BCVX7XN3U33HHYFDR6Z/action/author_attestation","sign_citation":"https://pith.science/pith/PEUKDK5BCVX7XN3U33HHYFDR6Z/action/citation_signature","submit_replication":"https://pith.science/pith/PEUKDK5BCVX7XN3U33HHYFDR6Z/action/replication_record"}},"created_at":"2026-07-05T11:34:52.016719+00:00","updated_at":"2026-07-05T11:34:52.016719+00:00"}