{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:HB65YYUGF6BSI3RCF2N24MFSEW","short_pith_number":"pith:HB65YYUG","canonical_record":{"source":{"id":"2405.09981","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-05-16T10:54:26Z","cross_cats_sorted":[],"title_canon_sha256":"f5f56d9130cb59cd5a6fb5c393f9f7b5c431dda9ffca7c4eec392c3429d8ccbb","abstract_canon_sha256":"50de7f352b89563362faf01d706025e11f24d017862f30e5e0ab494a143e52c9"},"schema_version":"1.0"},"canonical_sha256":"387ddc62862f83246e222e9bae30b225b74d001872b5f067ba365b422520473c","source":{"kind":"arxiv","id":"2405.09981","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.09981","created_at":"2026-07-05T08:19:51Z"},{"alias_kind":"arxiv_version","alias_value":"2405.09981v1","created_at":"2026-07-05T08:19:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.09981","created_at":"2026-07-05T08:19:51Z"},{"alias_kind":"pith_short_12","alias_value":"HB65YYUGF6BS","created_at":"2026-07-05T08:19:51Z"},{"alias_kind":"pith_short_16","alias_value":"HB65YYUGF6BSI3RC","created_at":"2026-07-05T08:19:51Z"},{"alias_kind":"pith_short_8","alias_value":"HB65YYUG","created_at":"2026-07-05T08:19:51Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:HB65YYUGF6BSI3RCF2N24MFSEW","target":"record","payload":{"canonical_record":{"source":{"id":"2405.09981","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-05-16T10:54:26Z","cross_cats_sorted":[],"title_canon_sha256":"f5f56d9130cb59cd5a6fb5c393f9f7b5c431dda9ffca7c4eec392c3429d8ccbb","abstract_canon_sha256":"50de7f352b89563362faf01d706025e11f24d017862f30e5e0ab494a143e52c9"},"schema_version":"1.0"},"canonical_sha256":"387ddc62862f83246e222e9bae30b225b74d001872b5f067ba365b422520473c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:19:51.454997Z","signature_b64":"yCPGI2eI6OapyhxCMXOZ0Zbfvmtrm1zl27jxhtU+gY+uI8B9nygUtUph2FmC9zLiofq0FGuD4g6YFlapWea1Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"387ddc62862f83246e222e9bae30b225b74d001872b5f067ba365b422520473c","last_reissued_at":"2026-07-05T08:19:51.454499Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:19:51.454499Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2405.09981","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:19:51Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0R1hQJJst+E1jKcZ1ADNmwOYBgaLyF8e6sB5aNDHTtLZrHYjuN/EG7ErceEqQBhxgFsbkZPcpzDB26QtSZDdCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T02:05:19.963861Z"},"content_sha256":"34df7f1e1be0778154ac93a284004692ab7002d6ab38a74f4e25079dce5a8f29","schema_version":"1.0","event_id":"sha256:34df7f1e1be0778154ac93a284004692ab7002d6ab38a74f4e25079dce5a8f29"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:HB65YYUGF6BSI3RCF2N24MFSEW","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Adversarial Robustness for Visual Grounding of Multimodal Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jiawang Bai, Kuofeng Gao, Shu-Tao Xia, Yang Bai, Yong Yang","submitted_at":"2024-05-16T10:54:26Z","abstract_excerpt":"Multi-modal Large Language Models (MLLMs) have recently achieved enhanced performance across various vision-language tasks including visual grounding capabilities. However, the adversarial robustness of visual grounding remains unexplored in MLLMs. To fill this gap, we use referring expression comprehension (REC) as an example task in visual grounding and propose three adversarial attack paradigms as follows. Firstly, untargeted adversarial attacks induce MLLMs to generate incorrect bounding boxes for each object. Besides, exclusive targeted adversarial attacks cause all generated outputs to t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.09981","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/2405.09981/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":"2026-07-05T08:19:51Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"E2UpoeadvkZomzy5nArVVQrapysGwV6/1ldVzaOr65ovCrOtnXOUEmyX6iU6iWjYu6ebKyjLEnRKA2MkrrtuCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T02:05:19.964368Z"},"content_sha256":"8742ae74a9c77c26c99dd7cac2f1111a15a0f4fa372e48f4b46774a6b2fd4f8d","schema_version":"1.0","event_id":"sha256:8742ae74a9c77c26c99dd7cac2f1111a15a0f4fa372e48f4b46774a6b2fd4f8d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/HB65YYUGF6BSI3RCF2N24MFSEW/bundle.json","state_url":"https://pith.science/pith/HB65YYUGF6BSI3RCF2N24MFSEW/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/HB65YYUGF6BSI3RCF2N24MFSEW/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":"2026-08-10T02:05:19Z","links":{"resolver":"https://pith.science/pith/HB65YYUGF6BSI3RCF2N24MFSEW","bundle":"https://pith.science/pith/HB65YYUGF6BSI3RCF2N24MFSEW/bundle.json","state":"https://pith.science/pith/HB65YYUGF6BSI3RCF2N24MFSEW/state.json","well_known_bundle":"https://pith.science/.well-known/pith/HB65YYUGF6BSI3RCF2N24MFSEW/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:HB65YYUGF6BSI3RCF2N24MFSEW","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":"50de7f352b89563362faf01d706025e11f24d017862f30e5e0ab494a143e52c9","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-05-16T10:54:26Z","title_canon_sha256":"f5f56d9130cb59cd5a6fb5c393f9f7b5c431dda9ffca7c4eec392c3429d8ccbb"},"schema_version":"1.0","source":{"id":"2405.09981","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.09981","created_at":"2026-07-05T08:19:51Z"},{"alias_kind":"arxiv_version","alias_value":"2405.09981v1","created_at":"2026-07-05T08:19:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.09981","created_at":"2026-07-05T08:19:51Z"},{"alias_kind":"pith_short_12","alias_value":"HB65YYUGF6BS","created_at":"2026-07-05T08:19:51Z"},{"alias_kind":"pith_short_16","alias_value":"HB65YYUGF6BSI3RC","created_at":"2026-07-05T08:19:51Z"},{"alias_kind":"pith_short_8","alias_value":"HB65YYUG","created_at":"2026-07-05T08:19:51Z"}],"graph_snapshots":[{"event_id":"sha256:8742ae74a9c77c26c99dd7cac2f1111a15a0f4fa372e48f4b46774a6b2fd4f8d","target":"graph","created_at":"2026-07-05T08:19:51Z","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/2405.09981/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multi-modal Large Language Models (MLLMs) have recently achieved enhanced performance across various vision-language tasks including visual grounding capabilities. However, the adversarial robustness of visual grounding remains unexplored in MLLMs. To fill this gap, we use referring expression comprehension (REC) as an example task in visual grounding and propose three adversarial attack paradigms as follows. Firstly, untargeted adversarial attacks induce MLLMs to generate incorrect bounding boxes for each object. Besides, exclusive targeted adversarial attacks cause all generated outputs to t","authors_text":"Jiawang Bai, Kuofeng Gao, Shu-Tao Xia, Yang Bai, Yong Yang","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-05-16T10:54:26Z","title":"Adversarial Robustness for Visual Grounding of Multimodal Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.09981","kind":"arxiv","version":1},"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":"sha256:34df7f1e1be0778154ac93a284004692ab7002d6ab38a74f4e25079dce5a8f29","target":"record","created_at":"2026-07-05T08:19:51Z","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":"50de7f352b89563362faf01d706025e11f24d017862f30e5e0ab494a143e52c9","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-05-16T10:54:26Z","title_canon_sha256":"f5f56d9130cb59cd5a6fb5c393f9f7b5c431dda9ffca7c4eec392c3429d8ccbb"},"schema_version":"1.0","source":{"id":"2405.09981","kind":"arxiv","version":1}},"canonical_sha256":"387ddc62862f83246e222e9bae30b225b74d001872b5f067ba365b422520473c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"387ddc62862f83246e222e9bae30b225b74d001872b5f067ba365b422520473c","first_computed_at":"2026-07-05T08:19:51.454499Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:19:51.454499Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"yCPGI2eI6OapyhxCMXOZ0Zbfvmtrm1zl27jxhtU+gY+uI8B9nygUtUph2FmC9zLiofq0FGuD4g6YFlapWea1Dg==","signature_status":"signed_v1","signed_at":"2026-07-05T08:19:51.454997Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.09981","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:34df7f1e1be0778154ac93a284004692ab7002d6ab38a74f4e25079dce5a8f29","sha256:8742ae74a9c77c26c99dd7cac2f1111a15a0f4fa372e48f4b46774a6b2fd4f8d"],"state_sha256":"ec1b39eb41e55bfad3f552d7a37fb4456934c9f6d7a5f4dc14ac979d250a8129"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3achkcbKIjEsYL2UIG1A+yUdLkUlRArF5S7HsbRw/QEzzQxQGW3N5pYBiZfO1KS0/cSTulaNNihdHPF3FotaDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T02:05:19.968635Z","bundle_sha256":"585b27f842ad99cf4e11f736999cb50844cd04c79285926751977a2b9618bc6a"}}