{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:VZMLLGAWQUW5Q3CRNEFGPHIF3K","short_pith_number":"pith:VZMLLGAW","schema_version":"1.0","canonical_sha256":"ae58b59816852dd86c51690a679d05daa289299a5e9e34bf7b86732fc9d054cd","source":{"kind":"arxiv","id":"2606.01503","version":1},"attestation_state":"computed","paper":{"title":"On the Limits of Token Reduction for Efficient Unified Vision Language Training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CV","authors_text":"Jingtao Li, Lingjuan Lv, Siyi Chen, Weiming Zhuang","submitted_at":"2026-05-31T23:59:12Z","abstract_excerpt":"Unified vision-language models (VLMs) integrate visual understanding and visual generation within a single autoregressive backbone, but their joint training is computationally expensive and largely overlooked from an efficiency perspective. In this work, we study the feasibility and limits of token-reduction-based acceleration for unified VLM training. Through a systematic analysis of layerwise attention allocation, we uncover a fundamental asymmetry: visual understanding exhibits substantial late-layer visual redundancy, whereas visual generation maintains persistent dependence on image token"},"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":"2606.01503","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-31T23:59:12Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"0d7963df82b5ed4618e757d9d35dd89f62d1676a6dfd3333afd7d30b9c9a7c18","abstract_canon_sha256":"601af94154317b76d74d0cc04eb134ffbb3e00546a67c90e84d6892edf8ec19e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:34.947841Z","signature_b64":"BBDN2+yuG0R1aIR7nBVy0zqogTOQBu4Va4bf6ikCIS3cB5c2IEa/3GMUe82K9zsPDj+msDN4evtZCZLOsyvgDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ae58b59816852dd86c51690a679d05daa289299a5e9e34bf7b86732fc9d054cd","last_reissued_at":"2026-06-02T02:04:34.947412Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:34.947412Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On the Limits of Token Reduction for Efficient Unified Vision Language Training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CV","authors_text":"Jingtao Li, Lingjuan Lv, Siyi Chen, Weiming Zhuang","submitted_at":"2026-05-31T23:59:12Z","abstract_excerpt":"Unified vision-language models (VLMs) integrate visual understanding and visual generation within a single autoregressive backbone, but their joint training is computationally expensive and largely overlooked from an efficiency perspective. In this work, we study the feasibility and limits of token-reduction-based acceleration for unified VLM training. Through a systematic analysis of layerwise attention allocation, we uncover a fundamental asymmetry: visual understanding exhibits substantial late-layer visual redundancy, whereas visual generation maintains persistent dependence on image token"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.01503","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/2606.01503/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":"2606.01503","created_at":"2026-06-02T02:04:34.947477+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.01503v1","created_at":"2026-06-02T02:04:34.947477+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.01503","created_at":"2026-06-02T02:04:34.947477+00:00"},{"alias_kind":"pith_short_12","alias_value":"VZMLLGAWQUW5","created_at":"2026-06-02T02:04:34.947477+00:00"},{"alias_kind":"pith_short_16","alias_value":"VZMLLGAWQUW5Q3CR","created_at":"2026-06-02T02:04:34.947477+00:00"},{"alias_kind":"pith_short_8","alias_value":"VZMLLGAW","created_at":"2026-06-02T02:04:34.947477+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/VZMLLGAWQUW5Q3CRNEFGPHIF3K","json":"https://pith.science/pith/VZMLLGAWQUW5Q3CRNEFGPHIF3K.json","graph_json":"https://pith.science/api/pith-number/VZMLLGAWQUW5Q3CRNEFGPHIF3K/graph.json","events_json":"https://pith.science/api/pith-number/VZMLLGAWQUW5Q3CRNEFGPHIF3K/events.json","paper":"https://pith.science/paper/VZMLLGAW"},"agent_actions":{"view_html":"https://pith.science/pith/VZMLLGAWQUW5Q3CRNEFGPHIF3K","download_json":"https://pith.science/pith/VZMLLGAWQUW5Q3CRNEFGPHIF3K.json","view_paper":"https://pith.science/paper/VZMLLGAW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.01503&json=true","fetch_graph":"https://pith.science/api/pith-number/VZMLLGAWQUW5Q3CRNEFGPHIF3K/graph.json","fetch_events":"https://pith.science/api/pith-number/VZMLLGAWQUW5Q3CRNEFGPHIF3K/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VZMLLGAWQUW5Q3CRNEFGPHIF3K/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VZMLLGAWQUW5Q3CRNEFGPHIF3K/action/storage_attestation","attest_author":"https://pith.science/pith/VZMLLGAWQUW5Q3CRNEFGPHIF3K/action/author_attestation","sign_citation":"https://pith.science/pith/VZMLLGAWQUW5Q3CRNEFGPHIF3K/action/citation_signature","submit_replication":"https://pith.science/pith/VZMLLGAWQUW5Q3CRNEFGPHIF3K/action/replication_record"}},"created_at":"2026-06-02T02:04:34.947477+00:00","updated_at":"2026-06-02T02:04:34.947477+00:00"}