{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:RLTYJFMEGC7H4G3TA5UWNDO75B","short_pith_number":"pith:RLTYJFME","canonical_record":{"source":{"id":"2605.12584","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-12T17:30:15Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"d30f037e7417ed359c3c399d7c54e871960abc3a0feee58bd01e9d1e7103d55b","abstract_canon_sha256":"d9019523ab3d6f8a94ae9cb00dfa8eb5512b2061462e18ed0a91c6e4c5ad91f7"},"schema_version":"1.0"},"canonical_sha256":"8ae784958430be7e1b730769668ddfe8476b9e229c29b85eee7321611ef60844","source":{"kind":"arxiv","id":"2605.12584","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.12584","created_at":"2026-05-18T03:10:01Z"},{"alias_kind":"arxiv_version","alias_value":"2605.12584v1","created_at":"2026-05-18T03:10:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.12584","created_at":"2026-05-18T03:10:01Z"},{"alias_kind":"pith_short_12","alias_value":"RLTYJFMEGC7H","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"RLTYJFMEGC7H4G3T","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"RLTYJFME","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:RLTYJFMEGC7H4G3TA5UWNDO75B","target":"record","payload":{"canonical_record":{"source":{"id":"2605.12584","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-12T17:30:15Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"d30f037e7417ed359c3c399d7c54e871960abc3a0feee58bd01e9d1e7103d55b","abstract_canon_sha256":"d9019523ab3d6f8a94ae9cb00dfa8eb5512b2061462e18ed0a91c6e4c5ad91f7"},"schema_version":"1.0"},"canonical_sha256":"8ae784958430be7e1b730769668ddfe8476b9e229c29b85eee7321611ef60844","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T03:10:01.391225Z","signature_b64":"ieLhzzo0L/3/JqG6Nf8NJtzkp4jA811Adb0BP7FSYtQMZToF0l2ZhncnPEsUHBQjRJaOPjdnu9LSdoOPMvlICg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8ae784958430be7e1b730769668ddfe8476b9e229c29b85eee7321611ef60844","last_reissued_at":"2026-05-18T03:10:01.390773Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T03:10:01.390773Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.12584","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":"2026-05-18T03:10:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CRKC7foOaCsXAriD8k2YyCkYiP+pbg/MItwQkS8/ieUrmLgceo/ilq5wuvU7TmFIyT4cyFu5ujzHzH2WG9XOAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T20:28:37.231199Z"},"content_sha256":"4a9cb9528db5b4a15f5bd191b1450f41a8addc2bf36e70ce873d25e5f1be5662","schema_version":"1.0","event_id":"sha256:4a9cb9528db5b4a15f5bd191b1450f41a8addc2bf36e70ce873d25e5f1be5662"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:RLTYJFMEGC7H4G3TA5UWNDO75B","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Towards Robust Federated Multimodal Graph Learning under Modality Heterogeneity","license":"http://creativecommons.org/licenses/by/4.0/","headline":"FedMPO recovers missing modalities in federated multimodal graphs using topology context and reliability-weighted aggregation.","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Guoren Wang, Haonan Wang, Hongchao Qin, Rong-Hua Li, Shumeng Li, Sirui Zhang, Xunkai Li, Zekai Chen","submitted_at":"2026-05-12T17:30:15Z","abstract_excerpt":"Recently, multimodal graph learning (MGL) has garnered significant attention for integrating diverse modality information and structured context to support various network applications. However, real-world graphs are often isolated due to data-sharing limitations across multiple parties, and their modalities are frequently incomplete. This highlights an urgent need to develop a robust federated approach. However, we find that existing methods remain insufficient. On the one hand, centralized MGL methods that handle missing modalities overlook the knowledge sharing and generalization in federat"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Extensive experiments on 3 tasks across 6 datasets demonstrate that FedMPO outperforms baselines, achieving performance gains of up to 4.10% and 5.65% in high-missing and non-IID settings.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That client-side topology-aware generation can reliably recover missing modalities from local graph context alone and that the reliability metric used for server aggregation accurately reflects true update quality without introducing new selection bias.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"FedMPO recovers missing modalities via topology-aware generation, filters noisy recoveries with missing-aware routing, and uses reliability-aware aggregation to achieve up to 5.65% gains over baselines in high-missing and non-IID federated graph settings.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"FedMPO recovers missing modalities in federated multimodal graphs using topology context and reliability-weighted aggregation.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"787a88a27b0619f3495a437bc1d75428ab5a75d42085ccf2786d6621471a337d"},"source":{"id":"2605.12584","kind":"arxiv","version":1},"verdict":{"id":"b6604d76-1f59-43eb-9b75-c78358746853","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T21:51:46.516152Z","strongest_claim":"Extensive experiments on 3 tasks across 6 datasets demonstrate that FedMPO outperforms baselines, achieving performance gains of up to 4.10% and 5.65% in high-missing and non-IID settings.","one_line_summary":"FedMPO recovers missing modalities via topology-aware generation, filters noisy recoveries with missing-aware routing, and uses reliability-aware aggregation to achieve up to 5.65% gains over baselines in high-missing and non-IID federated graph settings.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That client-side topology-aware generation can reliably recover missing modalities from local graph context alone and that the reliability metric used for server aggregation accurately reflects true update quality without introducing new selection bias.","pith_extraction_headline":"FedMPO recovers missing modalities in federated multimodal graphs using topology context and reliability-weighted aggregation."},"references":{"count":39,"sample":[{"doi":"","year":null,"title":"Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS) , series =","work_id":"5c76f482-e22e-4465-b6f3-eac01e5d083a","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Proceedings of Machine Learning and Systems , volume =","work_id":"5a1fb5ea-dd13-4492-9150-eceb58b93f44","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Proceedings of the 37th International Conference on Machine Learning (ICML) , series =","work_id":"d6cc51cb-cb6c-4f4a-bea6-27bafcd89b62","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Machine Learning and Knowledge Discovery in Databases","work_id":"4b4e153f-cfb4-4bc6-a764-4dffe2b653ca","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"and Le Nguyen, Phuong and Huynh, Trung Tin , title =","work_id":"341e4c64-3b6c-4bc8-ae60-bd9c2c94fcc3","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":39,"snapshot_sha256":"bbaf19ed3af6359278248c3861570f05aff4d23d78e0037a451cf8995d949973","internal_anchors":1},"formal_canon":{"evidence_count":2,"snapshot_sha256":"fe519fbf61e5d366963fb8585ffa56507bc97c5354feace3adda9f2bda0cec50"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"b6604d76-1f59-43eb-9b75-c78358746853"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T03:10:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZS2LBd+v8goFyPQYZ+S3SVR3KlRA2P6/8GxKGSWeGawrZ5356FQxf5N1KK95AEr0bqxInpxbhfQQJIrkSrZjBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T20:28:37.232233Z"},"content_sha256":"c4b313e792d7b037a88d83893ce20ea82ffd9540635ff5316128b58ee2c619ac","schema_version":"1.0","event_id":"sha256:c4b313e792d7b037a88d83893ce20ea82ffd9540635ff5316128b58ee2c619ac"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RLTYJFMEGC7H4G3TA5UWNDO75B/bundle.json","state_url":"https://pith.science/pith/RLTYJFMEGC7H4G3TA5UWNDO75B/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RLTYJFMEGC7H4G3TA5UWNDO75B/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":"2026-05-26T20:28:37Z","links":{"resolver":"https://pith.science/pith/RLTYJFMEGC7H4G3TA5UWNDO75B","bundle":"https://pith.science/pith/RLTYJFMEGC7H4G3TA5UWNDO75B/bundle.json","state":"https://pith.science/pith/RLTYJFMEGC7H4G3TA5UWNDO75B/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RLTYJFMEGC7H4G3TA5UWNDO75B/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:RLTYJFMEGC7H4G3TA5UWNDO75B","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":"d9019523ab3d6f8a94ae9cb00dfa8eb5512b2061462e18ed0a91c6e4c5ad91f7","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-12T17:30:15Z","title_canon_sha256":"d30f037e7417ed359c3c399d7c54e871960abc3a0feee58bd01e9d1e7103d55b"},"schema_version":"1.0","source":{"id":"2605.12584","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.12584","created_at":"2026-05-18T03:10:01Z"},{"alias_kind":"arxiv_version","alias_value":"2605.12584v1","created_at":"2026-05-18T03:10:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.12584","created_at":"2026-05-18T03:10:01Z"},{"alias_kind":"pith_short_12","alias_value":"RLTYJFMEGC7H","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"RLTYJFMEGC7H4G3T","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"RLTYJFME","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:c4b313e792d7b037a88d83893ce20ea82ffd9540635ff5316128b58ee2c619ac","target":"graph","created_at":"2026-05-18T03:10:01Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Extensive experiments on 3 tasks across 6 datasets demonstrate that FedMPO outperforms baselines, achieving performance gains of up to 4.10% and 5.65% in high-missing and non-IID settings."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That client-side topology-aware generation can reliably recover missing modalities from local graph context alone and that the reliability metric used for server aggregation accurately reflects true update quality without introducing new selection bias."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"FedMPO recovers missing modalities via topology-aware generation, filters noisy recoveries with missing-aware routing, and uses reliability-aware aggregation to achieve up to 5.65% gains over baselines in high-missing and non-IID federated graph settings."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"FedMPO recovers missing modalities in federated multimodal graphs using topology context and reliability-weighted aggregation."}],"snapshot_sha256":"787a88a27b0619f3495a437bc1d75428ab5a75d42085ccf2786d6621471a337d"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"fe519fbf61e5d366963fb8585ffa56507bc97c5354feace3adda9f2bda0cec50"},"paper":{"abstract_excerpt":"Recently, multimodal graph learning (MGL) has garnered significant attention for integrating diverse modality information and structured context to support various network applications. However, real-world graphs are often isolated due to data-sharing limitations across multiple parties, and their modalities are frequently incomplete. This highlights an urgent need to develop a robust federated approach. However, we find that existing methods remain insufficient. On the one hand, centralized MGL methods that handle missing modalities overlook the knowledge sharing and generalization in federat","authors_text":"Guoren Wang, Haonan Wang, Hongchao Qin, Rong-Hua Li, Shumeng Li, Sirui Zhang, Xunkai Li, Zekai Chen","cross_cats":["cs.AI"],"headline":"FedMPO recovers missing modalities in federated multimodal graphs using topology context and reliability-weighted aggregation.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-12T17:30:15Z","title":"Towards Robust Federated Multimodal Graph Learning under Modality Heterogeneity"},"references":{"count":39,"internal_anchors":1,"resolved_work":39,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS) , series =","work_id":"5c76f482-e22e-4465-b6f3-eac01e5d083a","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Proceedings of Machine Learning and Systems , volume =","work_id":"5a1fb5ea-dd13-4492-9150-eceb58b93f44","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Proceedings of the 37th International Conference on Machine Learning (ICML) , series =","work_id":"d6cc51cb-cb6c-4f4a-bea6-27bafcd89b62","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Machine Learning and Knowledge Discovery in Databases","work_id":"4b4e153f-cfb4-4bc6-a764-4dffe2b653ca","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"and Le Nguyen, Phuong and Huynh, Trung Tin , title =","work_id":"341e4c64-3b6c-4bc8-ae60-bd9c2c94fcc3","year":2024}],"snapshot_sha256":"bbaf19ed3af6359278248c3861570f05aff4d23d78e0037a451cf8995d949973"},"source":{"id":"2605.12584","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T21:51:46.516152Z","id":"b6604d76-1f59-43eb-9b75-c78358746853","model_set":{"reader":"grok-4.3"},"one_line_summary":"FedMPO recovers missing modalities via topology-aware generation, filters noisy recoveries with missing-aware routing, and uses reliability-aware aggregation to achieve up to 5.65% gains over baselines in high-missing and non-IID federated graph settings.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"FedMPO recovers missing modalities in federated multimodal graphs using topology context and reliability-weighted aggregation.","strongest_claim":"Extensive experiments on 3 tasks across 6 datasets demonstrate that FedMPO outperforms baselines, achieving performance gains of up to 4.10% and 5.65% in high-missing and non-IID settings.","weakest_assumption":"That client-side topology-aware generation can reliably recover missing modalities from local graph context alone and that the reliability metric used for server aggregation accurately reflects true update quality without introducing new selection bias."}},"verdict_id":"b6604d76-1f59-43eb-9b75-c78358746853"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:4a9cb9528db5b4a15f5bd191b1450f41a8addc2bf36e70ce873d25e5f1be5662","target":"record","created_at":"2026-05-18T03:10:01Z","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":"d9019523ab3d6f8a94ae9cb00dfa8eb5512b2061462e18ed0a91c6e4c5ad91f7","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-12T17:30:15Z","title_canon_sha256":"d30f037e7417ed359c3c399d7c54e871960abc3a0feee58bd01e9d1e7103d55b"},"schema_version":"1.0","source":{"id":"2605.12584","kind":"arxiv","version":1}},"canonical_sha256":"8ae784958430be7e1b730769668ddfe8476b9e229c29b85eee7321611ef60844","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8ae784958430be7e1b730769668ddfe8476b9e229c29b85eee7321611ef60844","first_computed_at":"2026-05-18T03:10:01.390773Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T03:10:01.390773Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ieLhzzo0L/3/JqG6Nf8NJtzkp4jA811Adb0BP7FSYtQMZToF0l2ZhncnPEsUHBQjRJaOPjdnu9LSdoOPMvlICg==","signature_status":"signed_v1","signed_at":"2026-05-18T03:10:01.391225Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.12584","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4a9cb9528db5b4a15f5bd191b1450f41a8addc2bf36e70ce873d25e5f1be5662","sha256:c4b313e792d7b037a88d83893ce20ea82ffd9540635ff5316128b58ee2c619ac"],"state_sha256":"44c470e4b594b5202d19877491c296016f2dafc6e4b760220385b6f33969a1cd"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4zR60KyqKC6n5hfFruWi4Pl2AKaPO1oa/wVawoTG6KIeGNZNxQZ6iEno9zZPCTyefS29CeoutTFxdkEOZEviBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-26T20:28:37.236327Z","bundle_sha256":"4ff6afdb29fb7507fde97acd14e9d825b4857dc0a582a1b06b8626cc9da4b731"}}