{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:SEPQ5USWJET5YNILIAWHIQDNSS","short_pith_number":"pith:SEPQ5USW","schema_version":"1.0","canonical_sha256":"911f0ed2564927dc350b402c74406d94b2d54cc0cc57d62492fbe1e9089a93b3","source":{"kind":"arxiv","id":"2406.11230","version":2},"attestation_state":"computed","paper":{"title":"Multimodal Needle in a Haystack: Benchmarking Long-Context Capability of Multimodal Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.CV"],"primary_cat":"cs.LG","authors_text":"Akshay Nambi, Haizhou Shi, Hao Wang, Hengyi Wang, Shiwei Tan, Tanuja Ganu, Tunyu Zhang, Weiyi Qin, Wenyuan Wang","submitted_at":"2024-06-17T05:54:06Z","abstract_excerpt":"Multimodal Large Language Models (MLLMs) have shown significant promise in various applications, leading to broad interest from researchers and practitioners alike. However, a comprehensive evaluation of their long-context capabilities remains underexplored. To address these gaps, we introduce the MultiModal Needle-in-a-haystack (MMNeedle) benchmark, specifically designed to assess the long-context capabilities of MLLMs. Besides multi-image input, we employ image stitching to further increase the input context length, and develop a protocol to automatically generate labels for sub-image level "},"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":"2406.11230","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-17T05:54:06Z","cross_cats_sorted":["cs.AI","cs.CL","cs.CV"],"title_canon_sha256":"28d89a539b26ed44fb76712fdb8781b159d9da505d6e64b931602ee9b2022a14","abstract_canon_sha256":"11c858c2460b8ad965875f168000c2bf9f5fbb856753c080bd0afd3840e59e09"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:12:17.141265Z","signature_b64":"+LPhpgC+MDFWZf/grsqPioiCyUKOYcdmUv1nAYXMpyFNuT76sHuL2Es7oEYrkCniaMhAyTF6HYjthV2gSNw2Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"911f0ed2564927dc350b402c74406d94b2d54cc0cc57d62492fbe1e9089a93b3","last_reissued_at":"2026-07-05T10:12:17.140769Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:12:17.140769Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Multimodal Needle in a Haystack: Benchmarking Long-Context Capability of Multimodal Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.CV"],"primary_cat":"cs.LG","authors_text":"Akshay Nambi, Haizhou Shi, Hao Wang, Hengyi Wang, Shiwei Tan, Tanuja Ganu, Tunyu Zhang, Weiyi Qin, Wenyuan Wang","submitted_at":"2024-06-17T05:54:06Z","abstract_excerpt":"Multimodal Large Language Models (MLLMs) have shown significant promise in various applications, leading to broad interest from researchers and practitioners alike. However, a comprehensive evaluation of their long-context capabilities remains underexplored. To address these gaps, we introduce the MultiModal Needle-in-a-haystack (MMNeedle) benchmark, specifically designed to assess the long-context capabilities of MLLMs. Besides multi-image input, we employ image stitching to further increase the input context length, and develop a protocol to automatically generate labels for sub-image level "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.11230","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/2406.11230/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":"2406.11230","created_at":"2026-07-05T10:12:17.140831+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.11230v2","created_at":"2026-07-05T10:12:17.140831+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.11230","created_at":"2026-07-05T10:12:17.140831+00:00"},{"alias_kind":"pith_short_12","alias_value":"SEPQ5USWJET5","created_at":"2026-07-05T10:12:17.140831+00:00"},{"alias_kind":"pith_short_16","alias_value":"SEPQ5USWJET5YNIL","created_at":"2026-07-05T10:12:17.140831+00:00"},{"alias_kind":"pith_short_8","alias_value":"SEPQ5USW","created_at":"2026-07-05T10:12:17.140831+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.14906","citing_title":"MemLens: Benchmarking Multimodal Long-Term Memory in Large Vision-Language Models","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2506.14493","citing_title":"LingoLoop Attack: Trapping MLLMs via Linguistic Context and State Entrapment into Endless Loops","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2602.14200","citing_title":"TS-Haystack: A Multi-Task Retrieval Benchmark for Long-Context Time-Series Reasoning","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13831","citing_title":"Training Long-Context Vision-Language Models Effectively with Generalization Beyond 128K Context","ref_index":6,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SEPQ5USWJET5YNILIAWHIQDNSS","json":"https://pith.science/pith/SEPQ5USWJET5YNILIAWHIQDNSS.json","graph_json":"https://pith.science/api/pith-number/SEPQ5USWJET5YNILIAWHIQDNSS/graph.json","events_json":"https://pith.science/api/pith-number/SEPQ5USWJET5YNILIAWHIQDNSS/events.json","paper":"https://pith.science/paper/SEPQ5USW"},"agent_actions":{"view_html":"https://pith.science/pith/SEPQ5USWJET5YNILIAWHIQDNSS","download_json":"https://pith.science/pith/SEPQ5USWJET5YNILIAWHIQDNSS.json","view_paper":"https://pith.science/paper/SEPQ5USW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.11230&json=true","fetch_graph":"https://pith.science/api/pith-number/SEPQ5USWJET5YNILIAWHIQDNSS/graph.json","fetch_events":"https://pith.science/api/pith-number/SEPQ5USWJET5YNILIAWHIQDNSS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SEPQ5USWJET5YNILIAWHIQDNSS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SEPQ5USWJET5YNILIAWHIQDNSS/action/storage_attestation","attest_author":"https://pith.science/pith/SEPQ5USWJET5YNILIAWHIQDNSS/action/author_attestation","sign_citation":"https://pith.science/pith/SEPQ5USWJET5YNILIAWHIQDNSS/action/citation_signature","submit_replication":"https://pith.science/pith/SEPQ5USWJET5YNILIAWHIQDNSS/action/replication_record"}},"created_at":"2026-07-05T10:12:17.140831+00:00","updated_at":"2026-07-05T10:12:17.140831+00:00"}