{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:DPXU56C5OQL2KO7HP6DETPPKYU","short_pith_number":"pith:DPXU56C5","schema_version":"1.0","canonical_sha256":"1bef4ef85d7417a53be77f8649bdeac52d821660543d11c14ac676cc0ffa4d9d","source":{"kind":"arxiv","id":"2404.18532","version":2},"attestation_state":"computed","paper":{"title":"MileBench: Benchmarking MLLMs in Long Context","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.LG"],"primary_cat":"cs.CL","authors_text":"Benyou Wang, Dingjie Song, Fei Yu, Guiming Hardy Chen, Shunian Chen, Xiang Wan","submitted_at":"2024-04-29T09:19:05Z","abstract_excerpt":"Despite the advancements and impressive performance of Multimodal Large Language Models (MLLMs) on benchmarks, their effectiveness in real-world, long-context, and multi-image tasks is unclear due to the benchmarks' limited scope. Existing benchmarks often focus on single-image and short-text samples, and when assessing multi-image tasks, they either limit the image count or focus on specific task (e.g time-series captioning), potentially obscuring the performance challenges of MLLMs. To address these limitations, we introduce MileBench, a pioneering benchmark designed to test the MultImodal L"},"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":"2404.18532","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-29T09:19:05Z","cross_cats_sorted":["cs.AI","cs.CV","cs.LG"],"title_canon_sha256":"79103ddbf4028d691c489ce354b42d87596dc64d5f2f6d570f2ed250ee7ace3c","abstract_canon_sha256":"83b4be5e8b5e102517ae34c50695cfd4383820c031726c67e5c8492302ecc968"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:19:24.514052Z","signature_b64":"ozgghJ0u1U83ArZ+yMdCQ1xiWv6m3lNDQmKDD9tLQnaEW7juBmuDXOAmYWFzi6BrddypKnewnpMtR2GPdlXpAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1bef4ef85d7417a53be77f8649bdeac52d821660543d11c14ac676cc0ffa4d9d","last_reissued_at":"2026-07-05T08:19:24.513562Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:19:24.513562Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MileBench: Benchmarking MLLMs in Long Context","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.LG"],"primary_cat":"cs.CL","authors_text":"Benyou Wang, Dingjie Song, Fei Yu, Guiming Hardy Chen, Shunian Chen, Xiang Wan","submitted_at":"2024-04-29T09:19:05Z","abstract_excerpt":"Despite the advancements and impressive performance of Multimodal Large Language Models (MLLMs) on benchmarks, their effectiveness in real-world, long-context, and multi-image tasks is unclear due to the benchmarks' limited scope. Existing benchmarks often focus on single-image and short-text samples, and when assessing multi-image tasks, they either limit the image count or focus on specific task (e.g time-series captioning), potentially obscuring the performance challenges of MLLMs. To address these limitations, we introduce MileBench, a pioneering benchmark designed to test the MultImodal L"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.18532","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/2404.18532/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":"2404.18532","created_at":"2026-07-05T08:19:24.513622+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.18532v2","created_at":"2026-07-05T08:19:24.513622+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.18532","created_at":"2026-07-05T08:19:24.513622+00:00"},{"alias_kind":"pith_short_12","alias_value":"DPXU56C5OQL2","created_at":"2026-07-05T08:19:24.513622+00:00"},{"alias_kind":"pith_short_16","alias_value":"DPXU56C5OQL2KO7H","created_at":"2026-07-05T08:19:24.513622+00:00"},{"alias_kind":"pith_short_8","alias_value":"DPXU56C5","created_at":"2026-07-05T08:19:24.513622+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26602","citing_title":"DiCoBench: Benchmarking Multi-Image Fine-Grained Perception via Differential and Commonality Visual Cues","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14906","citing_title":"MemLens: Benchmarking Multimodal Long-Term Memory in Large Vision-Language Models","ref_index":56,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15710","citing_title":"SMMBench: A Benchmark for Source-Distributed Multimodal Agent Memory","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27389","citing_title":"COHERENCE: Benchmarking Fine-Grained Image-Text Alignment in Interleaved Multimodal Contexts","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12703","citing_title":"MMCL-Bench: Multimodal Context Learning from Visual Rules, Procedures, and Evidence","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11591","citing_title":"Logit-Attention Divergence: Mitigating Position Bias in Multi-Image Retrieval via Attention-Guided Calibration","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27389","citing_title":"COHERENCE: Benchmarking Fine-Grained Image-Text Alignment in Interleaved Multimodal Contexts","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02262","citing_title":"WindowQuant: Mixed-Precision KV Cache Quantization based on Window-Level Similarity for VLMs Inference Optimization","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04075","citing_title":"RetentiveKV: State-Space Memory for Uncertainty-Aware Multimodal KV Cache Eviction","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05887","citing_title":"HybridKV: Hybrid KV Cache Compression for Efficient Multimodal Large Language Model Inference","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DPXU56C5OQL2KO7HP6DETPPKYU","json":"https://pith.science/pith/DPXU56C5OQL2KO7HP6DETPPKYU.json","graph_json":"https://pith.science/api/pith-number/DPXU56C5OQL2KO7HP6DETPPKYU/graph.json","events_json":"https://pith.science/api/pith-number/DPXU56C5OQL2KO7HP6DETPPKYU/events.json","paper":"https://pith.science/paper/DPXU56C5"},"agent_actions":{"view_html":"https://pith.science/pith/DPXU56C5OQL2KO7HP6DETPPKYU","download_json":"https://pith.science/pith/DPXU56C5OQL2KO7HP6DETPPKYU.json","view_paper":"https://pith.science/paper/DPXU56C5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.18532&json=true","fetch_graph":"https://pith.science/api/pith-number/DPXU56C5OQL2KO7HP6DETPPKYU/graph.json","fetch_events":"https://pith.science/api/pith-number/DPXU56C5OQL2KO7HP6DETPPKYU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DPXU56C5OQL2KO7HP6DETPPKYU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DPXU56C5OQL2KO7HP6DETPPKYU/action/storage_attestation","attest_author":"https://pith.science/pith/DPXU56C5OQL2KO7HP6DETPPKYU/action/author_attestation","sign_citation":"https://pith.science/pith/DPXU56C5OQL2KO7HP6DETPPKYU/action/citation_signature","submit_replication":"https://pith.science/pith/DPXU56C5OQL2KO7HP6DETPPKYU/action/replication_record"}},"created_at":"2026-07-05T08:19:24.513622+00:00","updated_at":"2026-07-05T08:19:24.513622+00:00"}