{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:K22FT5T2B3WTFQF36UQUFAHKFE","short_pith_number":"pith:K22FT5T2","schema_version":"1.0","canonical_sha256":"56b459f67a0eed32c0bbf5214280ea292b94c8a9a964b157e84119230a7381c1","source":{"kind":"arxiv","id":"2607.16326","version":1},"attestation_state":"computed","paper":{"title":"CRISP: Pre-LLM Yet Text-Driven Visual Token Pruning for Efficient LVLM Inference","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Baoquan Zhao, Juncen Guo, Mengyang Zhao, Rui Zhu, Wei Zhou, Xiaolei Chen, Xu Li, Yi Zheng, Yuxuan Liang, Zhe Liu","submitted_at":"2026-07-15T13:11:26Z","abstract_excerpt":"Large Vision-Language Models (LVLMs) typically require processing hundreds to thousands of visual tokens, leading to substantial inference overhead. Existing visual token pruning methods either operate before the LLM using text-agnostic heuristics or prune inside the LLM at the cost of efficiency and noisy cross-modal attention. To address these limitations, we propose CRISP, a pre-LLM yet text-driven visual token pruning framework that preserves both instruction-relevant evidence and essential scene context. CRISP works in a two-stage pipeline: Stage 1 first identifies text-aligned visual tok"},"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":"2607.16326","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-07-15T13:11:26Z","cross_cats_sorted":[],"title_canon_sha256":"26c2eb1fc64b3ed31d38ae5f12c0cddb11749478ffe2b3a1cc06bb0d4478ed29","abstract_canon_sha256":"99ad11b6ed6df93337877a291500e49a228e39907880ec74b5a1f36eea53b936"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-21T00:20:12.538477Z","signature_b64":"eYICHbJ6MREjfQbw+jczbgn+dhK9DWEnBp4v9Jwkef3edsIY8Li3xb3WyJyG1QAWHRuwNxA2mnFeAmsOluEkBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"56b459f67a0eed32c0bbf5214280ea292b94c8a9a964b157e84119230a7381c1","last_reissued_at":"2026-07-21T00:20:12.537561Z","signature_status":"signed_v1","first_computed_at":"2026-07-21T00:20:12.537561Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CRISP: Pre-LLM Yet Text-Driven Visual Token Pruning for Efficient LVLM Inference","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Baoquan Zhao, Juncen Guo, Mengyang Zhao, Rui Zhu, Wei Zhou, Xiaolei Chen, Xu Li, Yi Zheng, Yuxuan Liang, Zhe Liu","submitted_at":"2026-07-15T13:11:26Z","abstract_excerpt":"Large Vision-Language Models (LVLMs) typically require processing hundreds to thousands of visual tokens, leading to substantial inference overhead. Existing visual token pruning methods either operate before the LLM using text-agnostic heuristics or prune inside the LLM at the cost of efficiency and noisy cross-modal attention. To address these limitations, we propose CRISP, a pre-LLM yet text-driven visual token pruning framework that preserves both instruction-relevant evidence and essential scene context. CRISP works in a two-stage pipeline: Stage 1 first identifies text-aligned visual tok"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.16326","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/2607.16326/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":"2607.16326","created_at":"2026-07-21T00:20:12.538019+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.16326v1","created_at":"2026-07-21T00:20:12.538019+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.16326","created_at":"2026-07-21T00:20:12.538019+00:00"},{"alias_kind":"pith_short_12","alias_value":"K22FT5T2B3WT","created_at":"2026-07-21T00:20:12.538019+00:00"},{"alias_kind":"pith_short_16","alias_value":"K22FT5T2B3WTFQF3","created_at":"2026-07-21T00:20:12.538019+00:00"},{"alias_kind":"pith_short_8","alias_value":"K22FT5T2","created_at":"2026-07-21T00:20:12.538019+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/K22FT5T2B3WTFQF36UQUFAHKFE","json":"https://pith.science/pith/K22FT5T2B3WTFQF36UQUFAHKFE.json","graph_json":"https://pith.science/api/pith-number/K22FT5T2B3WTFQF36UQUFAHKFE/graph.json","events_json":"https://pith.science/api/pith-number/K22FT5T2B3WTFQF36UQUFAHKFE/events.json","paper":"https://pith.science/paper/K22FT5T2"},"agent_actions":{"view_html":"https://pith.science/pith/K22FT5T2B3WTFQF36UQUFAHKFE","download_json":"https://pith.science/pith/K22FT5T2B3WTFQF36UQUFAHKFE.json","view_paper":"https://pith.science/paper/K22FT5T2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.16326&json=true","fetch_graph":"https://pith.science/api/pith-number/K22FT5T2B3WTFQF36UQUFAHKFE/graph.json","fetch_events":"https://pith.science/api/pith-number/K22FT5T2B3WTFQF36UQUFAHKFE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/K22FT5T2B3WTFQF36UQUFAHKFE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/K22FT5T2B3WTFQF36UQUFAHKFE/action/storage_attestation","attest_author":"https://pith.science/pith/K22FT5T2B3WTFQF36UQUFAHKFE/action/author_attestation","sign_citation":"https://pith.science/pith/K22FT5T2B3WTFQF36UQUFAHKFE/action/citation_signature","submit_replication":"https://pith.science/pith/K22FT5T2B3WTFQF36UQUFAHKFE/action/replication_record"}},"created_at":"2026-07-21T00:20:12.538019+00:00","updated_at":"2026-07-21T00:20:12.538019+00:00"}