{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:YKIXVOEWAEQMUEYKAR3CWVFJWE","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":"5605af4ca86aea70a0f715ce537fb5ec5ca4511fc9d4e20d55628989ba632bfc","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-28T10:17:29Z","title_canon_sha256":"d5570c9b1a4da5f6179835ec3b4279f01cbb60944c5b7b96336fa453527a33d3"},"schema_version":"1.0","source":{"id":"2412.20105","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.20105","created_at":"2026-07-05T09:54:54Z"},{"alias_kind":"arxiv_version","alias_value":"2412.20105v1","created_at":"2026-07-05T09:54:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.20105","created_at":"2026-07-05T09:54:54Z"},{"alias_kind":"pith_short_12","alias_value":"YKIXVOEWAEQM","created_at":"2026-07-05T09:54:54Z"},{"alias_kind":"pith_short_16","alias_value":"YKIXVOEWAEQMUEYK","created_at":"2026-07-05T09:54:54Z"},{"alias_kind":"pith_short_8","alias_value":"YKIXVOEW","created_at":"2026-07-05T09:54:54Z"}],"graph_snapshots":[{"event_id":"sha256:165ef6a68643759e3e7ab08d616363f29cf87bc7c1ff636813f872ce3e4402e4","target":"graph","created_at":"2026-07-05T09:54:54Z","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/2412.20105/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multimodal large language models (MLLMs) enhance their perceptual capabilities by integrating visual and textual information. However, processing the massive number of visual tokens incurs a significant computational cost. Existing analysis of the MLLM attention mechanisms remains shallow, leading to coarse-grain token pruning strategies that fail to effectively balance speed and accuracy. In this paper, we conduct a comprehensive investigation of MLLM attention mechanisms with LLaVA. We find that numerous visual tokens and partial attention computations are redundant during the decoding proce","authors_text":"Haoji Hu, Jian Chen, Jiedong Zhuang, Lu Lu, Ming Dai, Qiang Liu, Rui Hu","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-28T10:17:29Z","title":"ST$^3$: Accelerating Multimodal Large Language Model by Spatial-Temporal Visual Token Trimming"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.20105","kind":"arxiv","version":1},"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:f86c3f1da319a8e4bf7b1c76fcbc4ccbc3b4cade65564cae98dfe1a83d5181c9","target":"record","created_at":"2026-07-05T09:54:54Z","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":"5605af4ca86aea70a0f715ce537fb5ec5ca4511fc9d4e20d55628989ba632bfc","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-28T10:17:29Z","title_canon_sha256":"d5570c9b1a4da5f6179835ec3b4279f01cbb60944c5b7b96336fa453527a33d3"},"schema_version":"1.0","source":{"id":"2412.20105","kind":"arxiv","version":1}},"canonical_sha256":"c2917ab8960120ca130a04762b54a9b12e9716f7ebcf5c75087930514d8ae5e7","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c2917ab8960120ca130a04762b54a9b12e9716f7ebcf5c75087930514d8ae5e7","first_computed_at":"2026-07-05T09:54:54.710671Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:54:54.710671Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Cqw1eO2Gug75ZJtCmsQlwnq/7mTXjmbnbO2xfIgZdvZvBDDQgVK1YOEuiHfvGkQ26pK1ezBtDO1EHcPTO3MZCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T09:54:54.711147Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.20105","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f86c3f1da319a8e4bf7b1c76fcbc4ccbc3b4cade65564cae98dfe1a83d5181c9","sha256:165ef6a68643759e3e7ab08d616363f29cf87bc7c1ff636813f872ce3e4402e4"],"state_sha256":"f80d34389c8412bee2f747842fa0bf54e2ff69b31f1901c7c1fddfcc60075c44"}