{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:3FOLZAGRZGLTF4PHDJBPMV7EGJ","short_pith_number":"pith:3FOLZAGR","schema_version":"1.0","canonical_sha256":"d95cbc80d1c99732f1e71a42f657e4327f5c91b69890038f5aa2be48d4c426ea","source":{"kind":"arxiv","id":"2605.15621","version":1},"attestation_state":"computed","paper":{"title":"LRCP: Low-Rank Compressibility Guided Visual Token Pruning for Efficient LVLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Feng Zhang, Hongyu Lu, Huanling Hu, Jiawei Li, Shikai Jiang, Tianjun Shi, Wenwei Jin, Yao Hu","submitted_at":"2026-05-15T05:09:15Z","abstract_excerpt":"Large vision-language models (LVLMs) achieve strong multimodal understanding, but their inference cost grows rapidly with the number of visual tokens, especially for high-resolution images and long videos. Existing attention-based methods estimate token importance from attention scores, which may introduce positional bias, while representation-based methods reduce visual redundancy based on feature relations or reconstruction errors, overlooking the global structure of the visual token set. In this paper, we revisit visual token compression from the perspective of low-rank compressibility. Acr"},"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":"2605.15621","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-15T05:09:15Z","cross_cats_sorted":[],"title_canon_sha256":"52cabd13215a9ca50d76188b9e402c48e5727e5d3021c3de31bc12ca65d2c846","abstract_canon_sha256":"7da479bdbd2780a56765ae0e28c0d216c346379349620a8efb41e36e2f2f3d4b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:01:08.651398Z","signature_b64":"UsiN8kK2Dwe2HZwF1PVcycNX+j9NXg44xI/brkvuyByp0Xvx+tZ2hYraAap3Xijbus7wC+XF9YkNfmvkraqHBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d95cbc80d1c99732f1e71a42f657e4327f5c91b69890038f5aa2be48d4c426ea","last_reissued_at":"2026-05-20T00:01:08.650670Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:01:08.650670Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LRCP: Low-Rank Compressibility Guided Visual Token Pruning for Efficient LVLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Feng Zhang, Hongyu Lu, Huanling Hu, Jiawei Li, Shikai Jiang, Tianjun Shi, Wenwei Jin, Yao Hu","submitted_at":"2026-05-15T05:09:15Z","abstract_excerpt":"Large vision-language models (LVLMs) achieve strong multimodal understanding, but their inference cost grows rapidly with the number of visual tokens, especially for high-resolution images and long videos. Existing attention-based methods estimate token importance from attention scores, which may introduce positional bias, while representation-based methods reduce visual redundancy based on feature relations or reconstruction errors, overlooking the global structure of the visual token set. In this paper, we revisit visual token compression from the perspective of low-rank compressibility. Acr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.15621","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/2605.15621/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-19T19:34:34.613443Z","status":"skipped","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T17:41:56.036301Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"b4ff70a4eec80360ff0ee8c0839e25ef0dd08613e5940f30099b1804ece648df"},"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":"2605.15621","created_at":"2026-05-20T00:01:08.650797+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.15621v1","created_at":"2026-05-20T00:01:08.650797+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.15621","created_at":"2026-05-20T00:01:08.650797+00:00"},{"alias_kind":"pith_short_12","alias_value":"3FOLZAGRZGLT","created_at":"2026-05-20T00:01:08.650797+00:00"},{"alias_kind":"pith_short_16","alias_value":"3FOLZAGRZGLTF4PH","created_at":"2026-05-20T00:01:08.650797+00:00"},{"alias_kind":"pith_short_8","alias_value":"3FOLZAGR","created_at":"2026-05-20T00:01:08.650797+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/3FOLZAGRZGLTF4PHDJBPMV7EGJ","json":"https://pith.science/pith/3FOLZAGRZGLTF4PHDJBPMV7EGJ.json","graph_json":"https://pith.science/api/pith-number/3FOLZAGRZGLTF4PHDJBPMV7EGJ/graph.json","events_json":"https://pith.science/api/pith-number/3FOLZAGRZGLTF4PHDJBPMV7EGJ/events.json","paper":"https://pith.science/paper/3FOLZAGR"},"agent_actions":{"view_html":"https://pith.science/pith/3FOLZAGRZGLTF4PHDJBPMV7EGJ","download_json":"https://pith.science/pith/3FOLZAGRZGLTF4PHDJBPMV7EGJ.json","view_paper":"https://pith.science/paper/3FOLZAGR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.15621&json=true","fetch_graph":"https://pith.science/api/pith-number/3FOLZAGRZGLTF4PHDJBPMV7EGJ/graph.json","fetch_events":"https://pith.science/api/pith-number/3FOLZAGRZGLTF4PHDJBPMV7EGJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3FOLZAGRZGLTF4PHDJBPMV7EGJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3FOLZAGRZGLTF4PHDJBPMV7EGJ/action/storage_attestation","attest_author":"https://pith.science/pith/3FOLZAGRZGLTF4PHDJBPMV7EGJ/action/author_attestation","sign_citation":"https://pith.science/pith/3FOLZAGRZGLTF4PHDJBPMV7EGJ/action/citation_signature","submit_replication":"https://pith.science/pith/3FOLZAGRZGLTF4PHDJBPMV7EGJ/action/replication_record"}},"created_at":"2026-05-20T00:01:08.650797+00:00","updated_at":"2026-05-20T00:01:08.650797+00:00"}