{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:YMPNMKCQWTXYWQTKZMHZQYGJPO","short_pith_number":"pith:YMPNMKCQ","schema_version":"1.0","canonical_sha256":"c31ed62850b4ef8b426acb0f9860c97b851ec739f2f6ef5d639c24b8b9e92586","source":{"kind":"arxiv","id":"2412.06458","version":2},"attestation_state":"computed","paper":{"title":"Pruning All-Rounder: Rethinking and Improving Inference Efficiency for Large Vision Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ji Ma, Lin Yuanbo Wu, Mengyang Sun, Peng Wang, Wei Suo, Yanning Zhang","submitted_at":"2024-12-09T13:02:35Z","abstract_excerpt":"Although Large Vision-Language Models (LVLMs) have achieved impressive results, their high computational costs pose a significant barrier to wide application. To enhance inference efficiency, most existing approaches can be categorized as parameter-dependent or token-dependent strategies to reduce computational demands. However, parameter-dependent methods require retraining LVLMs to recover performance while token-dependent strategies struggle to consistently select the most relevant tokens. In this paper, we systematically analyze the above challenges and provide a series of valuable insight"},"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":"2412.06458","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-09T13:02:35Z","cross_cats_sorted":[],"title_canon_sha256":"7b971fdfd8277069e1e5ba7fdce34ad9e69c545ce3990090bb77dc00a3c3dd87","abstract_canon_sha256":"513c05a8af49ea97932ec4df0423fc92c24bf00945ef03e7a6b51c1bfd8a063a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:45:54.714764Z","signature_b64":"DxXfIDxU9+8pZMLvV5BSuPDkUcIBsUHOPIUwY8KSRhwKgR9ENRMP426UsB4TXSsHt4t5UPpJRBJ+hrrbeCm4AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c31ed62850b4ef8b426acb0f9860c97b851ec739f2f6ef5d639c24b8b9e92586","last_reissued_at":"2026-07-05T11:45:54.714200Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:45:54.714200Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Pruning All-Rounder: Rethinking and Improving Inference Efficiency for Large Vision Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ji Ma, Lin Yuanbo Wu, Mengyang Sun, Peng Wang, Wei Suo, Yanning Zhang","submitted_at":"2024-12-09T13:02:35Z","abstract_excerpt":"Although Large Vision-Language Models (LVLMs) have achieved impressive results, their high computational costs pose a significant barrier to wide application. To enhance inference efficiency, most existing approaches can be categorized as parameter-dependent or token-dependent strategies to reduce computational demands. However, parameter-dependent methods require retraining LVLMs to recover performance while token-dependent strategies struggle to consistently select the most relevant tokens. In this paper, we systematically analyze the above challenges and provide a series of valuable insight"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.06458","kind":"arxiv","version":2},"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/2412.06458/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":"2412.06458","created_at":"2026-07-05T11:45:54.714259+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.06458v2","created_at":"2026-07-05T11:45:54.714259+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.06458","created_at":"2026-07-05T11:45:54.714259+00:00"},{"alias_kind":"pith_short_12","alias_value":"YMPNMKCQWTXY","created_at":"2026-07-05T11:45:54.714259+00:00"},{"alias_kind":"pith_short_16","alias_value":"YMPNMKCQWTXYWQTK","created_at":"2026-07-05T11:45:54.714259+00:00"},{"alias_kind":"pith_short_8","alias_value":"YMPNMKCQ","created_at":"2026-07-05T11:45:54.714259+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.01236","citing_title":"Mitigating Information Loss under High Pruning Rates for Efficient Large Vision Language Models","ref_index":44,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YMPNMKCQWTXYWQTKZMHZQYGJPO","json":"https://pith.science/pith/YMPNMKCQWTXYWQTKZMHZQYGJPO.json","graph_json":"https://pith.science/api/pith-number/YMPNMKCQWTXYWQTKZMHZQYGJPO/graph.json","events_json":"https://pith.science/api/pith-number/YMPNMKCQWTXYWQTKZMHZQYGJPO/events.json","paper":"https://pith.science/paper/YMPNMKCQ"},"agent_actions":{"view_html":"https://pith.science/pith/YMPNMKCQWTXYWQTKZMHZQYGJPO","download_json":"https://pith.science/pith/YMPNMKCQWTXYWQTKZMHZQYGJPO.json","view_paper":"https://pith.science/paper/YMPNMKCQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.06458&json=true","fetch_graph":"https://pith.science/api/pith-number/YMPNMKCQWTXYWQTKZMHZQYGJPO/graph.json","fetch_events":"https://pith.science/api/pith-number/YMPNMKCQWTXYWQTKZMHZQYGJPO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YMPNMKCQWTXYWQTKZMHZQYGJPO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YMPNMKCQWTXYWQTKZMHZQYGJPO/action/storage_attestation","attest_author":"https://pith.science/pith/YMPNMKCQWTXYWQTKZMHZQYGJPO/action/author_attestation","sign_citation":"https://pith.science/pith/YMPNMKCQWTXYWQTKZMHZQYGJPO/action/citation_signature","submit_replication":"https://pith.science/pith/YMPNMKCQWTXYWQTKZMHZQYGJPO/action/replication_record"}},"created_at":"2026-07-05T11:45:54.714259+00:00","updated_at":"2026-07-05T11:45:54.714259+00:00"}