{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:SHYIADY4F7QL2OC7PXYE2STLR4","short_pith_number":"pith:SHYIADY4","schema_version":"1.0","canonical_sha256":"91f0800f1c2fe0bd385f7df04d4a6b8f006d13cdb9c37d664db8c23d2e269fc8","source":{"kind":"arxiv","id":"2607.27952","version":1},"attestation_state":"computed","paper":{"title":"LAST: The Last Query Token Guides Visual Token Pruning for Edge-Cloud Collaborative MLLM Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Chris Xing Tian, Feng Yang, Haoliang Li, Howard Leung, Keyang Zhang, Rongqun Lin, Shiqi Wang, Xiandong Meng, Xinrui Ju","submitted_at":"2026-07-30T09:59:25Z","abstract_excerpt":"Multimodal foundation models are reshaping edge-cloud visual intelligence from task-specific feature pipelines into token-based interfaces, where edge devices encode visual inputs into tokens for a general-purpose cloud MLLM. However, dense visual-token sequences increase cloud-side inference costs. Existing pruning methods mainly target centralized inference: vision-driven methods can operate before cloud execution but are typically query-agnostic, whereas query-guided methods often rely on internal states of the target MLLM and cannot determine token relevance before transmission. Compact gu"},"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":"2607.27952","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-07-30T09:59:25Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"800b2f56756d261539785bfc5e8ef96c4a1821ed2c966717efc9d8e638251846","abstract_canon_sha256":"82f5a3e761650c1a028db1585f91d8b7da656159c2503558f26a8c7f01d68382"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"91f0800f1c2fe0bd385f7df04d4a6b8f006d13cdb9c37d664db8c23d2e269fc8","last_reissued_at":"2026-07-31T01:35:00.865215Z","signature_status":"unsigned_v0","first_computed_at":"2026-07-31T01:35:00.865215Z"},"graph_snapshot":{"paper":{"title":"LAST: The Last Query Token Guides Visual Token Pruning for Edge-Cloud Collaborative MLLM Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Chris Xing Tian, Feng Yang, Haoliang Li, Howard Leung, Keyang Zhang, Rongqun Lin, Shiqi Wang, Xiandong Meng, Xinrui Ju","submitted_at":"2026-07-30T09:59:25Z","abstract_excerpt":"Multimodal foundation models are reshaping edge-cloud visual intelligence from task-specific feature pipelines into token-based interfaces, where edge devices encode visual inputs into tokens for a general-purpose cloud MLLM. However, dense visual-token sequences increase cloud-side inference costs. Existing pruning methods mainly target centralized inference: vision-driven methods can operate before cloud execution but are typically query-agnostic, whereas query-guided methods often rely on internal states of the target MLLM and cannot determine token relevance before transmission. Compact gu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.27952","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/2607.27952/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":"2607.27952","created_at":"2026-07-31T01:35:00.868569+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.27952v1","created_at":"2026-07-31T01:35:00.868569+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.27952","created_at":"2026-07-31T01:35:00.868569+00:00"},{"alias_kind":"pith_short_12","alias_value":"SHYIADY4F7QL","created_at":"2026-07-31T01:35:00.868569+00:00"},{"alias_kind":"pith_short_16","alias_value":"SHYIADY4F7QL2OC7","created_at":"2026-07-31T01:35:00.868569+00:00"},{"alias_kind":"pith_short_8","alias_value":"SHYIADY4","created_at":"2026-07-31T01:35:00.868569+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/SHYIADY4F7QL2OC7PXYE2STLR4","json":"https://pith.science/pith/SHYIADY4F7QL2OC7PXYE2STLR4.json","graph_json":"https://pith.science/api/pith-number/SHYIADY4F7QL2OC7PXYE2STLR4/graph.json","events_json":"https://pith.science/api/pith-number/SHYIADY4F7QL2OC7PXYE2STLR4/events.json","paper":"https://pith.science/paper/SHYIADY4"},"agent_actions":{"view_html":"https://pith.science/pith/SHYIADY4F7QL2OC7PXYE2STLR4","download_json":"https://pith.science/pith/SHYIADY4F7QL2OC7PXYE2STLR4.json","view_paper":"https://pith.science/paper/SHYIADY4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.27952&json=true","fetch_graph":"https://pith.science/api/pith-number/SHYIADY4F7QL2OC7PXYE2STLR4/graph.json","fetch_events":"https://pith.science/api/pith-number/SHYIADY4F7QL2OC7PXYE2STLR4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SHYIADY4F7QL2OC7PXYE2STLR4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SHYIADY4F7QL2OC7PXYE2STLR4/action/storage_attestation","attest_author":"https://pith.science/pith/SHYIADY4F7QL2OC7PXYE2STLR4/action/author_attestation","sign_citation":"https://pith.science/pith/SHYIADY4F7QL2OC7PXYE2STLR4/action/citation_signature","submit_replication":"https://pith.science/pith/SHYIADY4F7QL2OC7PXYE2STLR4/action/replication_record"}},"created_at":"2026-07-31T01:35:00.868569+00:00","updated_at":"2026-07-31T01:35:00.868569+00:00"}