{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:7XRZZWS2SRDZFQFLWTTUUSZGOC","short_pith_number":"pith:7XRZZWS2","schema_version":"1.0","canonical_sha256":"fde39cda5a944792c0abb4e74a4b2670abbec7b72c671cd49d736c51817bddcb","source":{"kind":"arxiv","id":"2503.14827","version":1},"attestation_state":"computed","paper":{"title":"MMDT: Decoding the Trustworthiness and Safety of Multimodal Foundation Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.CL","authors_text":"Alexander Xiong, Andy Zhou, Bo Li, Chejian Xu, Chengquan Guo, Chenhui Zhang, Chulin Xie, Dan Hendrycks, Dawn Song, Francesco Pinto, Jeffrey Ziwei Tan, Jiawei Zhang, Lingzhi Yuan, Mintong Kang, Peiyang Xu, Xuandong Zhao, Yi Zeng, Yu Gai, Yujin Potter, Zhaorun Chen, Zhen Xiang, Zhun Wang, Zhuowen Yuan, Zidi Xiong, Zinan Lin","submitted_at":"2025-03-19T01:59:44Z","abstract_excerpt":"Multimodal foundation models (MMFMs) play a crucial role in various applications, including autonomous driving, healthcare, and virtual assistants. However, several studies have revealed vulnerabilities in these models, such as generating unsafe content by text-to-image models. Existing benchmarks on multimodal models either predominantly assess the helpfulness of these models, or only focus on limited perspectives such as fairness and privacy. In this paper, we present the first unified platform, MMDT (Multimodal DecodingTrust), designed to provide a comprehensive safety and trustworthiness e"},"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":"2503.14827","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-19T01:59:44Z","cross_cats_sorted":["cs.AI","cs.CR"],"title_canon_sha256":"d8f079d4559d515a704a0bfef98eab2867ba64efc743ae3fc57ca3bc6739189d","abstract_canon_sha256":"63b20c788a067aa76cf2212a2418025f4af9ab3c9788df9ce775c4509f21b3b0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:34:25.256881Z","signature_b64":"URtWyJ51E4I94LlxMVhh9ear+WgB98oTrewfDSba43yBL4Dc7x9A0JQhCcYmBWCBYqbs0t4UZYYp7uu6iRHaCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fde39cda5a944792c0abb4e74a4b2670abbec7b72c671cd49d736c51817bddcb","last_reissued_at":"2026-07-05T10:34:25.256028Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:34:25.256028Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MMDT: Decoding the Trustworthiness and Safety of Multimodal Foundation Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.CL","authors_text":"Alexander Xiong, Andy Zhou, Bo Li, Chejian Xu, Chengquan Guo, Chenhui Zhang, Chulin Xie, Dan Hendrycks, Dawn Song, Francesco Pinto, Jeffrey Ziwei Tan, Jiawei Zhang, Lingzhi Yuan, Mintong Kang, Peiyang Xu, Xuandong Zhao, Yi Zeng, Yu Gai, Yujin Potter, Zhaorun Chen, Zhen Xiang, Zhun Wang, Zhuowen Yuan, Zidi Xiong, Zinan Lin","submitted_at":"2025-03-19T01:59:44Z","abstract_excerpt":"Multimodal foundation models (MMFMs) play a crucial role in various applications, including autonomous driving, healthcare, and virtual assistants. However, several studies have revealed vulnerabilities in these models, such as generating unsafe content by text-to-image models. Existing benchmarks on multimodal models either predominantly assess the helpfulness of these models, or only focus on limited perspectives such as fairness and privacy. In this paper, we present the first unified platform, MMDT (Multimodal DecodingTrust), designed to provide a comprehensive safety and trustworthiness e"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.14827","kind":"arxiv","version":1},"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/2503.14827/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":"2503.14827","created_at":"2026-07-05T10:34:25.256138+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.14827v1","created_at":"2026-07-05T10:34:25.256138+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.14827","created_at":"2026-07-05T10:34:25.256138+00:00"},{"alias_kind":"pith_short_12","alias_value":"7XRZZWS2SRDZ","created_at":"2026-07-05T10:34:25.256138+00:00"},{"alias_kind":"pith_short_16","alias_value":"7XRZZWS2SRDZFQFL","created_at":"2026-07-05T10:34:25.256138+00:00"},{"alias_kind":"pith_short_8","alias_value":"7XRZZWS2","created_at":"2026-07-05T10:34:25.256138+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7XRZZWS2SRDZFQFLWTTUUSZGOC","json":"https://pith.science/pith/7XRZZWS2SRDZFQFLWTTUUSZGOC.json","graph_json":"https://pith.science/api/pith-number/7XRZZWS2SRDZFQFLWTTUUSZGOC/graph.json","events_json":"https://pith.science/api/pith-number/7XRZZWS2SRDZFQFLWTTUUSZGOC/events.json","paper":"https://pith.science/paper/7XRZZWS2"},"agent_actions":{"view_html":"https://pith.science/pith/7XRZZWS2SRDZFQFLWTTUUSZGOC","download_json":"https://pith.science/pith/7XRZZWS2SRDZFQFLWTTUUSZGOC.json","view_paper":"https://pith.science/paper/7XRZZWS2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.14827&json=true","fetch_graph":"https://pith.science/api/pith-number/7XRZZWS2SRDZFQFLWTTUUSZGOC/graph.json","fetch_events":"https://pith.science/api/pith-number/7XRZZWS2SRDZFQFLWTTUUSZGOC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7XRZZWS2SRDZFQFLWTTUUSZGOC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7XRZZWS2SRDZFQFLWTTUUSZGOC/action/storage_attestation","attest_author":"https://pith.science/pith/7XRZZWS2SRDZFQFLWTTUUSZGOC/action/author_attestation","sign_citation":"https://pith.science/pith/7XRZZWS2SRDZFQFLWTTUUSZGOC/action/citation_signature","submit_replication":"https://pith.science/pith/7XRZZWS2SRDZFQFLWTTUUSZGOC/action/replication_record"}},"created_at":"2026-07-05T10:34:25.256138+00:00","updated_at":"2026-07-05T10:34:25.256138+00:00"}