{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:OVVJNLFFWX6MU5B3QWJQMUUOD7","short_pith_number":"pith:OVVJNLFF","schema_version":"1.0","canonical_sha256":"756a96aca5b5fcca743b859306528e1fce9b6a51ab334ce045b3b37d48f6008b","source":{"kind":"arxiv","id":"2407.01509","version":5},"attestation_state":"computed","paper":{"title":"MIA-Bench: Towards Better Instruction Following Evaluation of Multimodal LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Hanrong Ye, Jean-Philippe Fauconnier, Peter Grasch, Yinfei Yang, Yusu Qian, Zhe Gan","submitted_at":"2024-07-01T17:53:35Z","abstract_excerpt":"We introduce MIA-Bench, a new benchmark designed to evaluate multimodal large language models (MLLMs) on their ability to strictly adhere to complex instructions. Our benchmark comprises a diverse set of 400 image-prompt pairs, each crafted to challenge the models' compliance with layered instructions in generating accurate responses that satisfy specific requested patterns. Evaluation results from a wide array of state-of-the-art MLLMs reveal significant variations in performance, highlighting areas for improvement in instruction fidelity. Additionally, we create extra training data and explo"},"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":"2407.01509","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-07-01T17:53:35Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"784146118c756944401f72de017854405ed04d189f927931d84d9b2f22d4bd40","abstract_canon_sha256":"8e60131523b87061752ebc334e0007629675188daec45cb5cbfa686c8d7e6451"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:35:34.150628Z","signature_b64":"iO0GxhVr9S5ARINhZh2fG48+PrdU4fDtkrlbnRgRFBMxXUh3rlY0YHfcPAMT0GBGW4OTuviLFT84Bc79hqa3Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"756a96aca5b5fcca743b859306528e1fce9b6a51ab334ce045b3b37d48f6008b","last_reissued_at":"2026-07-05T10:35:34.149659Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:35:34.149659Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MIA-Bench: Towards Better Instruction Following Evaluation of Multimodal LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Hanrong Ye, Jean-Philippe Fauconnier, Peter Grasch, Yinfei Yang, Yusu Qian, Zhe Gan","submitted_at":"2024-07-01T17:53:35Z","abstract_excerpt":"We introduce MIA-Bench, a new benchmark designed to evaluate multimodal large language models (MLLMs) on their ability to strictly adhere to complex instructions. Our benchmark comprises a diverse set of 400 image-prompt pairs, each crafted to challenge the models' compliance with layered instructions in generating accurate responses that satisfy specific requested patterns. Evaluation results from a wide array of state-of-the-art MLLMs reveal significant variations in performance, highlighting areas for improvement in instruction fidelity. Additionally, we create extra training data and explo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.01509","kind":"arxiv","version":5},"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/2407.01509/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":"2407.01509","created_at":"2026-07-05T10:35:34.149799+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.01509v5","created_at":"2026-07-05T10:35:34.149799+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.01509","created_at":"2026-07-05T10:35:34.149799+00:00"},{"alias_kind":"pith_short_12","alias_value":"OVVJNLFFWX6M","created_at":"2026-07-05T10:35:34.149799+00:00"},{"alias_kind":"pith_short_16","alias_value":"OVVJNLFFWX6MU5B3","created_at":"2026-07-05T10:35:34.149799+00:00"},{"alias_kind":"pith_short_8","alias_value":"OVVJNLFF","created_at":"2026-07-05T10:35:34.149799+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.12703","citing_title":"MMCL-Bench: Multimodal Context Learning from Visual Rules, Procedures, and Evidence","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18320","citing_title":"EVE: Verifiable Self-Evolution of MLLMs via Executable Visual Transformations","ref_index":31,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OVVJNLFFWX6MU5B3QWJQMUUOD7","json":"https://pith.science/pith/OVVJNLFFWX6MU5B3QWJQMUUOD7.json","graph_json":"https://pith.science/api/pith-number/OVVJNLFFWX6MU5B3QWJQMUUOD7/graph.json","events_json":"https://pith.science/api/pith-number/OVVJNLFFWX6MU5B3QWJQMUUOD7/events.json","paper":"https://pith.science/paper/OVVJNLFF"},"agent_actions":{"view_html":"https://pith.science/pith/OVVJNLFFWX6MU5B3QWJQMUUOD7","download_json":"https://pith.science/pith/OVVJNLFFWX6MU5B3QWJQMUUOD7.json","view_paper":"https://pith.science/paper/OVVJNLFF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.01509&json=true","fetch_graph":"https://pith.science/api/pith-number/OVVJNLFFWX6MU5B3QWJQMUUOD7/graph.json","fetch_events":"https://pith.science/api/pith-number/OVVJNLFFWX6MU5B3QWJQMUUOD7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OVVJNLFFWX6MU5B3QWJQMUUOD7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OVVJNLFFWX6MU5B3QWJQMUUOD7/action/storage_attestation","attest_author":"https://pith.science/pith/OVVJNLFFWX6MU5B3QWJQMUUOD7/action/author_attestation","sign_citation":"https://pith.science/pith/OVVJNLFFWX6MU5B3QWJQMUUOD7/action/citation_signature","submit_replication":"https://pith.science/pith/OVVJNLFFWX6MU5B3QWJQMUUOD7/action/replication_record"}},"created_at":"2026-07-05T10:35:34.149799+00:00","updated_at":"2026-07-05T10:35:34.149799+00:00"}