{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:CLEULMRNFOE4QUZJF6L7P56JNR","short_pith_number":"pith:CLEULMRN","canonical_record":{"source":{"id":"2412.00556","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-11-30T18:54:32Z","cross_cats_sorted":[],"title_canon_sha256":"7f05341b7f6344f280cf0108eb1d6bf2aedea6d674d4bfd1226a9878a9d7d0af","abstract_canon_sha256":"71fcbd5fc5130f898f10206ebf30f7933a17fe390c384456d183c029ab8d7e72"},"schema_version":"1.0"},"canonical_sha256":"12c945b22d2b89c853292f97f7f7c96c4a295c3218327e29b4f30d634d408d0e","source":{"kind":"arxiv","id":"2412.00556","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.00556","created_at":"2026-07-05T09:45:59Z"},{"alias_kind":"arxiv_version","alias_value":"2412.00556v2","created_at":"2026-07-05T09:45:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.00556","created_at":"2026-07-05T09:45:59Z"},{"alias_kind":"pith_short_12","alias_value":"CLEULMRNFOE4","created_at":"2026-07-05T09:45:59Z"},{"alias_kind":"pith_short_16","alias_value":"CLEULMRNFOE4QUZJ","created_at":"2026-07-05T09:45:59Z"},{"alias_kind":"pith_short_8","alias_value":"CLEULMRN","created_at":"2026-07-05T09:45:59Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:CLEULMRNFOE4QUZJF6L7P56JNR","target":"record","payload":{"canonical_record":{"source":{"id":"2412.00556","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-11-30T18:54:32Z","cross_cats_sorted":[],"title_canon_sha256":"7f05341b7f6344f280cf0108eb1d6bf2aedea6d674d4bfd1226a9878a9d7d0af","abstract_canon_sha256":"71fcbd5fc5130f898f10206ebf30f7933a17fe390c384456d183c029ab8d7e72"},"schema_version":"1.0"},"canonical_sha256":"12c945b22d2b89c853292f97f7f7c96c4a295c3218327e29b4f30d634d408d0e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:45:59.183654Z","signature_b64":"+3l6Up+UB3pmR9w7C/sxI8YSP99FYobHONy1gPPWb/3JFt8YcNcZQqf3WqnZIFh3vFa8/opkEsI1DHy+IqlVDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"12c945b22d2b89c853292f97f7f7c96c4a295c3218327e29b4f30d634d408d0e","last_reissued_at":"2026-07-05T09:45:59.183185Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:45:59.183185Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.00556","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":"2026-07-05T09:45:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xbAMJVIXlzNn02MttiCkFVtJiRfum877RL0Wdjcg0vzuqm8CClAD/egLGc1RTVgQzurBt0NYh2vBsLJc9leaCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-13T19:18:47.706890Z"},"content_sha256":"f75bf049b5e25b0d838696901c9eb032917ad5c212050120299e7137523c1692","schema_version":"1.0","event_id":"sha256:f75bf049b5e25b0d838696901c9eb032917ad5c212050120299e7137523c1692"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:CLEULMRNFOE4QUZJF6L7P56JNR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Accelerating Multimodal Large Language Models by Searching Optimal Vision Token Reduction","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dimitris N. Metaxas, Felix Juefei-Xu, Licheng Yu, Miao Liu, Mingfu Liang, Ning Zhang, Shiyu Zhao, Xiaofang Wang, Xide Xia, Zhenting Wang","submitted_at":"2024-11-30T18:54:32Z","abstract_excerpt":"Prevailing Multimodal Large Language Models (MLLMs) encode the input image(s) as vision tokens and feed them into the language backbone, similar to how Large Language Models (LLMs) process the text tokens. However, the number of vision tokens increases quadratically as the image resolutions, leading to huge computational costs. In this paper, we consider improving MLLM's efficiency from two scenarios, (I) Reducing computational cost without degrading the performance. (II) Improving the performance with given budgets. We start with our main finding that the ranking of each vision token sorted b"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.00556","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/2412.00556/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":"2026-07-05T09:45:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"WabWTPyZGT0PSxyceLBqna886wKw0UDoXXXBUGp6rroSTcUbAUAyOhHPrLiTRu5jZm0ey/M+Cl/DruU4xlUeBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-13T19:18:47.707577Z"},"content_sha256":"f6733333de9ac911b359411e4e832a02c39bafdcc4da6aa44be334a72a7f89a0","schema_version":"1.0","event_id":"sha256:f6733333de9ac911b359411e4e832a02c39bafdcc4da6aa44be334a72a7f89a0"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/CLEULMRNFOE4QUZJF6L7P56JNR/bundle.json","state_url":"https://pith.science/pith/CLEULMRNFOE4QUZJF6L7P56JNR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/CLEULMRNFOE4QUZJF6L7P56JNR/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":"2026-08-13T19:18:47Z","links":{"resolver":"https://pith.science/pith/CLEULMRNFOE4QUZJF6L7P56JNR","bundle":"https://pith.science/pith/CLEULMRNFOE4QUZJF6L7P56JNR/bundle.json","state":"https://pith.science/pith/CLEULMRNFOE4QUZJF6L7P56JNR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/CLEULMRNFOE4QUZJF6L7P56JNR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:CLEULMRNFOE4QUZJF6L7P56JNR","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":"71fcbd5fc5130f898f10206ebf30f7933a17fe390c384456d183c029ab8d7e72","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-11-30T18:54:32Z","title_canon_sha256":"7f05341b7f6344f280cf0108eb1d6bf2aedea6d674d4bfd1226a9878a9d7d0af"},"schema_version":"1.0","source":{"id":"2412.00556","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.00556","created_at":"2026-07-05T09:45:59Z"},{"alias_kind":"arxiv_version","alias_value":"2412.00556v2","created_at":"2026-07-05T09:45:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.00556","created_at":"2026-07-05T09:45:59Z"},{"alias_kind":"pith_short_12","alias_value":"CLEULMRNFOE4","created_at":"2026-07-05T09:45:59Z"},{"alias_kind":"pith_short_16","alias_value":"CLEULMRNFOE4QUZJ","created_at":"2026-07-05T09:45:59Z"},{"alias_kind":"pith_short_8","alias_value":"CLEULMRN","created_at":"2026-07-05T09:45:59Z"}],"graph_snapshots":[{"event_id":"sha256:f6733333de9ac911b359411e4e832a02c39bafdcc4da6aa44be334a72a7f89a0","target":"graph","created_at":"2026-07-05T09:45:59Z","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/2412.00556/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Prevailing Multimodal Large Language Models (MLLMs) encode the input image(s) as vision tokens and feed them into the language backbone, similar to how Large Language Models (LLMs) process the text tokens. However, the number of vision tokens increases quadratically as the image resolutions, leading to huge computational costs. In this paper, we consider improving MLLM's efficiency from two scenarios, (I) Reducing computational cost without degrading the performance. (II) Improving the performance with given budgets. We start with our main finding that the ranking of each vision token sorted b","authors_text":"Dimitris N. Metaxas, Felix Juefei-Xu, Licheng Yu, Miao Liu, Mingfu Liang, Ning Zhang, Shiyu Zhao, Xiaofang Wang, Xide Xia, Zhenting Wang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-11-30T18:54:32Z","title":"Accelerating Multimodal Large Language Models by Searching Optimal Vision Token Reduction"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.00556","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:f75bf049b5e25b0d838696901c9eb032917ad5c212050120299e7137523c1692","target":"record","created_at":"2026-07-05T09:45:59Z","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":"71fcbd5fc5130f898f10206ebf30f7933a17fe390c384456d183c029ab8d7e72","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-11-30T18:54:32Z","title_canon_sha256":"7f05341b7f6344f280cf0108eb1d6bf2aedea6d674d4bfd1226a9878a9d7d0af"},"schema_version":"1.0","source":{"id":"2412.00556","kind":"arxiv","version":2}},"canonical_sha256":"12c945b22d2b89c853292f97f7f7c96c4a295c3218327e29b4f30d634d408d0e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"12c945b22d2b89c853292f97f7f7c96c4a295c3218327e29b4f30d634d408d0e","first_computed_at":"2026-07-05T09:45:59.183185Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:45:59.183185Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"+3l6Up+UB3pmR9w7C/sxI8YSP99FYobHONy1gPPWb/3JFt8YcNcZQqf3WqnZIFh3vFa8/opkEsI1DHy+IqlVDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T09:45:59.183654Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.00556","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f75bf049b5e25b0d838696901c9eb032917ad5c212050120299e7137523c1692","sha256:f6733333de9ac911b359411e4e832a02c39bafdcc4da6aa44be334a72a7f89a0"],"state_sha256":"f9d46313d50575601f90ea840fd33135896e688821aaecae414d3d069aea8c1a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fmCGe8ikHBWFgoeMOqw93sn415ZgR0TBLlE/XlqaVmTiVbsWAllSH6Y17j5zTkHlL/zuFhFb82D1Qx8NyKVoDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-13T19:18:47.768981Z","bundle_sha256":"ccdab71b35f7e209d8462dafca0bdfa54b745d1ad1c9ff7ca8f9da36c1e0c8c8"}}