{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:Z2QMCJLCZWB2V66V2EKKRQHHPE","short_pith_number":"pith:Z2QMCJLC","canonical_record":{"source":{"id":"2412.01818","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-02T18:57:40Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"0344ecc6c44ffb602d0707a16e4e848ff75007aee0aa99f8876b403ce7133a31","abstract_canon_sha256":"067565c7f64791952612cc1479d90f15b67b928fac3e8ae2e00da4c2872ab4c7"},"schema_version":"1.0"},"canonical_sha256":"cea0c12562cd83aafbd5d114a8c0e7791ce865320a5d150d4a9603ddfb44d810","source":{"kind":"arxiv","id":"2412.01818","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.01818","created_at":"2026-07-05T11:01:29Z"},{"alias_kind":"arxiv_version","alias_value":"2412.01818v2","created_at":"2026-07-05T11:01:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.01818","created_at":"2026-07-05T11:01:29Z"},{"alias_kind":"pith_short_12","alias_value":"Z2QMCJLCZWB2","created_at":"2026-07-05T11:01:29Z"},{"alias_kind":"pith_short_16","alias_value":"Z2QMCJLCZWB2V66V","created_at":"2026-07-05T11:01:29Z"},{"alias_kind":"pith_short_8","alias_value":"Z2QMCJLC","created_at":"2026-07-05T11:01:29Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:Z2QMCJLCZWB2V66V2EKKRQHHPE","target":"record","payload":{"canonical_record":{"source":{"id":"2412.01818","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-02T18:57:40Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"0344ecc6c44ffb602d0707a16e4e848ff75007aee0aa99f8876b403ce7133a31","abstract_canon_sha256":"067565c7f64791952612cc1479d90f15b67b928fac3e8ae2e00da4c2872ab4c7"},"schema_version":"1.0"},"canonical_sha256":"cea0c12562cd83aafbd5d114a8c0e7791ce865320a5d150d4a9603ddfb44d810","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:01:29.491711Z","signature_b64":"C9nPKnL8ytMXKyM5UOH83LSklAswUdRPgm5qOCZdmbFlQYADR4C4iuTzHlG7MQ2ydvYFvXdpZ9daQjMFKILhCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cea0c12562cd83aafbd5d114a8c0e7791ce865320a5d150d4a9603ddfb44d810","last_reissued_at":"2026-07-05T11:01:29.491217Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:01:29.491217Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.01818","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:01:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OXAdlxlBaLBteOmmGgzlks65WcNzS5zFRxVNQTXe5gVEk2aVFD+/UhIOufF15arJTop/cvLDkbpX9m7hme+ECA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T20:47:50.513981Z"},"content_sha256":"00d7dda1c3f5d35e9adae4c4c6f5a4ba93696a86f9219d613ce9355db6ee3b28","schema_version":"1.0","event_id":"sha256:00d7dda1c3f5d35e9adae4c4c6f5a4ba93696a86f9219d613ce9355db6ee3b28"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:Z2QMCJLCZWB2V66V2EKKRQHHPE","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Beyond Text-Visual Attention: Exploiting Visual Cues for Effective Token Pruning in VLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Aosong Cheng, Jiajun Cao, Ming Lu, Qi She, Qizhe Zhang, Renrui Zhang, Shanghang Zhang, Shaobo Guo, Zhiyong Zhuo","submitted_at":"2024-12-02T18:57:40Z","abstract_excerpt":"Large vision-language models (LVLMs) generally contain significantly more visual tokens than their textual counterparts, resulting in a considerable computational burden. Recent efforts have been made to tackle this issue by pruning visual tokens early within the language model. Most existing works use attention scores between text and visual tokens to assess the importance of visual tokens. However, in this study, we first analyze the text-visual attention in the language model and find that this score is not an ideal indicator for token pruning. Based on the analysis, We propose VisPruner, a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.01818","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/2412.01818/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:01:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"1dJnYsHEeSFF7HtCIi1JyaA6vIE/e+ukregMhcWEf3Blc3ATPaI1DH7jj5xt2bBge+H/1SqL66kjmX6XL9MHDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T20:47:50.514495Z"},"content_sha256":"b8974fcbeafe14b4b3efc247e3bf62a93dd0877a18816804a20455db38297326","schema_version":"1.0","event_id":"sha256:b8974fcbeafe14b4b3efc247e3bf62a93dd0877a18816804a20455db38297326"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/Z2QMCJLCZWB2V66V2EKKRQHHPE/bundle.json","state_url":"https://pith.science/pith/Z2QMCJLCZWB2V66V2EKKRQHHPE/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/Z2QMCJLCZWB2V66V2EKKRQHHPE/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-08T20:47:50Z","links":{"resolver":"https://pith.science/pith/Z2QMCJLCZWB2V66V2EKKRQHHPE","bundle":"https://pith.science/pith/Z2QMCJLCZWB2V66V2EKKRQHHPE/bundle.json","state":"https://pith.science/pith/Z2QMCJLCZWB2V66V2EKKRQHHPE/state.json","well_known_bundle":"https://pith.science/.well-known/pith/Z2QMCJLCZWB2V66V2EKKRQHHPE/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:Z2QMCJLCZWB2V66V2EKKRQHHPE","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"067565c7f64791952612cc1479d90f15b67b928fac3e8ae2e00da4c2872ab4c7","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-02T18:57:40Z","title_canon_sha256":"0344ecc6c44ffb602d0707a16e4e848ff75007aee0aa99f8876b403ce7133a31"},"schema_version":"1.0","source":{"id":"2412.01818","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.01818","created_at":"2026-07-05T11:01:29Z"},{"alias_kind":"arxiv_version","alias_value":"2412.01818v2","created_at":"2026-07-05T11:01:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.01818","created_at":"2026-07-05T11:01:29Z"},{"alias_kind":"pith_short_12","alias_value":"Z2QMCJLCZWB2","created_at":"2026-07-05T11:01:29Z"},{"alias_kind":"pith_short_16","alias_value":"Z2QMCJLCZWB2V66V","created_at":"2026-07-05T11:01:29Z"},{"alias_kind":"pith_short_8","alias_value":"Z2QMCJLC","created_at":"2026-07-05T11:01:29Z"}],"graph_snapshots":[{"event_id":"sha256:b8974fcbeafe14b4b3efc247e3bf62a93dd0877a18816804a20455db38297326","target":"graph","created_at":"2026-07-05T11:01:29Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2412.01818/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large vision-language models (LVLMs) generally contain significantly more visual tokens than their textual counterparts, resulting in a considerable computational burden. Recent efforts have been made to tackle this issue by pruning visual tokens early within the language model. Most existing works use attention scores between text and visual tokens to assess the importance of visual tokens. However, in this study, we first analyze the text-visual attention in the language model and find that this score is not an ideal indicator for token pruning. Based on the analysis, We propose VisPruner, a","authors_text":"Aosong Cheng, Jiajun Cao, Ming Lu, Qi She, Qizhe Zhang, Renrui Zhang, Shanghang Zhang, Shaobo Guo, Zhiyong Zhuo","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-02T18:57:40Z","title":"Beyond Text-Visual Attention: Exploiting Visual Cues for Effective Token Pruning in VLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.01818","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:00d7dda1c3f5d35e9adae4c4c6f5a4ba93696a86f9219d613ce9355db6ee3b28","target":"record","created_at":"2026-07-05T11:01:29Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"067565c7f64791952612cc1479d90f15b67b928fac3e8ae2e00da4c2872ab4c7","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-02T18:57:40Z","title_canon_sha256":"0344ecc6c44ffb602d0707a16e4e848ff75007aee0aa99f8876b403ce7133a31"},"schema_version":"1.0","source":{"id":"2412.01818","kind":"arxiv","version":2}},"canonical_sha256":"cea0c12562cd83aafbd5d114a8c0e7791ce865320a5d150d4a9603ddfb44d810","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"cea0c12562cd83aafbd5d114a8c0e7791ce865320a5d150d4a9603ddfb44d810","first_computed_at":"2026-07-05T11:01:29.491217Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:01:29.491217Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"C9nPKnL8ytMXKyM5UOH83LSklAswUdRPgm5qOCZdmbFlQYADR4C4iuTzHlG7MQ2ydvYFvXdpZ9daQjMFKILhCA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:01:29.491711Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.01818","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:00d7dda1c3f5d35e9adae4c4c6f5a4ba93696a86f9219d613ce9355db6ee3b28","sha256:b8974fcbeafe14b4b3efc247e3bf62a93dd0877a18816804a20455db38297326"],"state_sha256":"e241b43ce34b335bfdaf980b9dbe9bf24da53f22bf69c650cfc615fe8e2df20a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ILv4rqyOv0mbncmEA8rAmadGkVoVXxYu3Y1VfDtrwA3vrHvFocZbr0v1U05QiYG2MdVMVtfAeqFK8GJAeMljDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T20:47:50.517855Z","bundle_sha256":"ca03d4cc010a7ccdd50bbdf47f984de9ec116422dd5afbf7de8d8d4f654371d2"}}