{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2012:PII5Q5WV6SLT4E5T25LCZRGA2A","short_pith_number":"pith:PII5Q5WV","canonical_record":{"source":{"id":"1205.2618","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2012-05-09T18:25:09Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"a2ae5ee8235f28fc19ef50476f72eb3c83c25f5b36bb5546e91b38660f177512","abstract_canon_sha256":"892b0d523a721ff71fc0c2bb6c9327782bb694b8845a702c7f431065a4578439"},"schema_version":"1.0"},"canonical_sha256":"7a11d876d5f4973e13b3d7562cc4c0d024ea0678b23167d1feb3501d273c707a","source":{"kind":"arxiv","id":"1205.2618","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1205.2618","created_at":"2026-05-18T03:55:47Z"},{"alias_kind":"arxiv_version","alias_value":"1205.2618v1","created_at":"2026-05-18T03:55:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1205.2618","created_at":"2026-05-18T03:55:47Z"},{"alias_kind":"pith_short_12","alias_value":"PII5Q5WV6SLT","created_at":"2026-05-18T12:27:18Z"},{"alias_kind":"pith_short_16","alias_value":"PII5Q5WV6SLT4E5T","created_at":"2026-05-18T12:27:18Z"},{"alias_kind":"pith_short_8","alias_value":"PII5Q5WV","created_at":"2026-05-18T12:27:18Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2012:PII5Q5WV6SLT4E5T25LCZRGA2A","target":"record","payload":{"canonical_record":{"source":{"id":"1205.2618","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2012-05-09T18:25:09Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"a2ae5ee8235f28fc19ef50476f72eb3c83c25f5b36bb5546e91b38660f177512","abstract_canon_sha256":"892b0d523a721ff71fc0c2bb6c9327782bb694b8845a702c7f431065a4578439"},"schema_version":"1.0"},"canonical_sha256":"7a11d876d5f4973e13b3d7562cc4c0d024ea0678b23167d1feb3501d273c707a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T03:55:47.887260Z","signature_b64":"4w/Q/bjqxMAFcY5Y0G1sghsyhsRFNMvX6wIZUHVm/PDu+zdSOcX9QqoVIGl0paiQmUILNkZgcFpKg+IvdZ1SBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7a11d876d5f4973e13b3d7562cc4c0d024ea0678b23167d1feb3501d273c707a","last_reissued_at":"2026-05-18T03:55:47.886750Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T03:55:47.886750Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1205.2618","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":"2026-05-18T03:55:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"25A5jVaz49BnrWnA+zc/UXife1yVn4lbb24TlVyNlyYVHun20Nxz6vDYXqc90CqL3L1jQlPeXt4bcx01coK4BA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T16:40:56.835194Z"},"content_sha256":"3800ad311e853618ee9c60cfaaa2805267af96b257e12e1c646e6e09b683f7bb","schema_version":"1.0","event_id":"sha256:3800ad311e853618ee9c60cfaaa2805267af96b257e12e1c646e6e09b683f7bb"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2012:PII5Q5WV6SLT4E5T25LCZRGA2A","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"BPR: Bayesian Personalized Ranking from Implicit Feedback","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Bayesian Personalized Ranking derives an optimization criterion that directly targets ranking from implicit feedback.","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.IR","authors_text":"Christoph Freudenthaler, Lars Schmidt-Thieme, Steffen Rendle, Zeno Gantner","submitted_at":"2012-05-09T18:25:09Z","abstract_excerpt":"Item recommendation is the task of predicting a personalized ranking on a set of items (e.g. websites, movies, products). In this paper, we investigate the most common scenario with implicit feedback (e.g. clicks, purchases). There are many methods for item recommendation from implicit feedback like matrix factorization (MF) or adaptive knearest-neighbor (kNN). Even though these methods are designed for the item prediction task of personalized ranking, none of them is directly optimized for ranking. In this paper we present a generic optimization criterion BPR-Opt for personalized ranking that"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Our experiments indicate that for the task of personalized ranking our optimization method outperforms the standard learning techniques for MF and kNN.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumption that maximizing the BPR-Opt posterior (derived from a Bayesian analysis of pairwise preferences) produces better ranking performance than standard losses when applied to matrix factorization and kNN models.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"BPR-Opt is a maximum posterior estimator for personalized ranking from implicit feedback that, when used to train matrix factorization and adaptive kNN, outperforms standard learning techniques on ranking tasks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Bayesian Personalized Ranking derives an optimization criterion that directly targets ranking from implicit feedback.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"1fade7c78a7a13333bbfe52711a515728b063f294de5b9ddff19d9e122b420fc"},"source":{"id":"1205.2618","kind":"arxiv","version":1},"verdict":{"id":"11bcfaf3-975b-4df5-8f6c-1fe2fb34c349","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-13T22:59:32.218731Z","strongest_claim":"Our experiments indicate that for the task of personalized ranking our optimization method outperforms the standard learning techniques for MF and kNN.","one_line_summary":"BPR-Opt is a maximum posterior estimator for personalized ranking from implicit feedback that, when used to train matrix factorization and adaptive kNN, outperforms standard learning techniques on ranking tasks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumption that maximizing the BPR-Opt posterior (derived from a Bayesian analysis of pairwise preferences) produces better ranking performance than standard losses when applied to matrix factorization and kNN models.","pith_extraction_headline":"Bayesian Personalized Ranking derives an optimization criterion that directly targets ranking from implicit feedback."},"references":{"count":15,"sample":[{"doi":"","year":2005,"title":"C. Burges, T. Shaked, E. Renshaw, A. Lazier, M. Deeds, N. Hamilton, and G. Hullender. Learn- ing to rank using gradient descent. In ICML ’05: Proceedings of the 22nd international con- ference on Mach","work_id":"d027aa54-31c9-4eeb-8152-9d36137b5e05","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2004,"title":"M. Deshpande and G. Karypis. Item-based top-n recommendation algorithms. ACM Transactions on Information Systems. Springer-Verlag , 22/1, 2004","work_id":"ece2b173-a33f-480e-9292-3e6fe0740d3a","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2004,"title":"A. Herschtal and B. Raskutti. Optimising area under the roc curve using gradient descent. In ICML ’04: Proceedings of the twenty-first inter- national conference on Machine learning, page 49, New York,","work_id":"16241d65-2ae4-49a1-a313-d9aebc422590","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2004,"title":"T. Hofmann. Latent semantic models for collabo- rative filtering. ACM Trans. Inf. Syst. , 22(1):89– 115, 2004","work_id":"30cb0301-626d-4e11-8402-99f2239c3cf6","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2008,"title":"Y. Hu, Y. Koren, and C. Volinsky. Collaborative filtering for implicit feedback datasets. In IEEE International Conference on Data Mining (ICDM 2008), pages 263–272, 2008","work_id":"9e666f2d-7918-4a97-89ba-fd0fb71ea901","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":15,"snapshot_sha256":"e84876f435cf0296916d908c4a2ced6cd66a1d998113a001750031c130ceb9eb","internal_anchors":0},"formal_canon":{"evidence_count":1,"snapshot_sha256":"9fd969aed509e5fd2f0c5f4adec5bf6488d5ed4653ac8733824a800633680c21"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"11bcfaf3-975b-4df5-8f6c-1fe2fb34c349"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T03:55:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TzxJWAv7p0+kDAcX6Z3bCoyf52tTV1BZXXQuClzwqim1s/Vrt80IRWsD5XwIqlUdPxf5E9DqEzb/bwWoGzLtBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T16:40:56.836536Z"},"content_sha256":"f721c8a1bbcff204f344617174aa71bae5e7ce4aac1a9a864bcec09a0b0d0a3c","schema_version":"1.0","event_id":"sha256:f721c8a1bbcff204f344617174aa71bae5e7ce4aac1a9a864bcec09a0b0d0a3c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/PII5Q5WV6SLT4E5T25LCZRGA2A/bundle.json","state_url":"https://pith.science/pith/PII5Q5WV6SLT4E5T25LCZRGA2A/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/PII5Q5WV6SLT4E5T25LCZRGA2A/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":"2026-05-25T16:40:56Z","links":{"resolver":"https://pith.science/pith/PII5Q5WV6SLT4E5T25LCZRGA2A","bundle":"https://pith.science/pith/PII5Q5WV6SLT4E5T25LCZRGA2A/bundle.json","state":"https://pith.science/pith/PII5Q5WV6SLT4E5T25LCZRGA2A/state.json","well_known_bundle":"https://pith.science/.well-known/pith/PII5Q5WV6SLT4E5T25LCZRGA2A/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2012:PII5Q5WV6SLT4E5T25LCZRGA2A","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":"892b0d523a721ff71fc0c2bb6c9327782bb694b8845a702c7f431065a4578439","cross_cats_sorted":["cs.LG","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2012-05-09T18:25:09Z","title_canon_sha256":"a2ae5ee8235f28fc19ef50476f72eb3c83c25f5b36bb5546e91b38660f177512"},"schema_version":"1.0","source":{"id":"1205.2618","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1205.2618","created_at":"2026-05-18T03:55:47Z"},{"alias_kind":"arxiv_version","alias_value":"1205.2618v1","created_at":"2026-05-18T03:55:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1205.2618","created_at":"2026-05-18T03:55:47Z"},{"alias_kind":"pith_short_12","alias_value":"PII5Q5WV6SLT","created_at":"2026-05-18T12:27:18Z"},{"alias_kind":"pith_short_16","alias_value":"PII5Q5WV6SLT4E5T","created_at":"2026-05-18T12:27:18Z"},{"alias_kind":"pith_short_8","alias_value":"PII5Q5WV","created_at":"2026-05-18T12:27:18Z"}],"graph_snapshots":[{"event_id":"sha256:f721c8a1bbcff204f344617174aa71bae5e7ce4aac1a9a864bcec09a0b0d0a3c","target":"graph","created_at":"2026-05-18T03:55:47Z","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":"Our experiments indicate that for the task of personalized ranking our optimization method outperforms the standard learning techniques for MF and kNN."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The assumption that maximizing the BPR-Opt posterior (derived from a Bayesian analysis of pairwise preferences) produces better ranking performance than standard losses when applied to matrix factorization and kNN models."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"BPR-Opt is a maximum posterior estimator for personalized ranking from implicit feedback that, when used to train matrix factorization and adaptive kNN, outperforms standard learning techniques on ranking tasks."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Bayesian Personalized Ranking derives an optimization criterion that directly targets ranking from implicit feedback."}],"snapshot_sha256":"1fade7c78a7a13333bbfe52711a515728b063f294de5b9ddff19d9e122b420fc"},"formal_canon":{"evidence_count":1,"snapshot_sha256":"9fd969aed509e5fd2f0c5f4adec5bf6488d5ed4653ac8733824a800633680c21"},"paper":{"abstract_excerpt":"Item recommendation is the task of predicting a personalized ranking on a set of items (e.g. websites, movies, products). In this paper, we investigate the most common scenario with implicit feedback (e.g. clicks, purchases). There are many methods for item recommendation from implicit feedback like matrix factorization (MF) or adaptive knearest-neighbor (kNN). Even though these methods are designed for the item prediction task of personalized ranking, none of them is directly optimized for ranking. In this paper we present a generic optimization criterion BPR-Opt for personalized ranking that","authors_text":"Christoph Freudenthaler, Lars Schmidt-Thieme, Steffen Rendle, Zeno Gantner","cross_cats":["cs.LG","stat.ML"],"headline":"Bayesian Personalized Ranking derives an optimization criterion that directly targets ranking from implicit feedback.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2012-05-09T18:25:09Z","title":"BPR: Bayesian Personalized Ranking from Implicit Feedback"},"references":{"count":15,"internal_anchors":0,"resolved_work":15,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"C. Burges, T. Shaked, E. Renshaw, A. Lazier, M. Deeds, N. Hamilton, and G. Hullender. Learn- ing to rank using gradient descent. In ICML ’05: Proceedings of the 22nd international con- ference on Mach","work_id":"d027aa54-31c9-4eeb-8152-9d36137b5e05","year":2005},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"M. Deshpande and G. Karypis. Item-based top-n recommendation algorithms. ACM Transactions on Information Systems. Springer-Verlag , 22/1, 2004","work_id":"ece2b173-a33f-480e-9292-3e6fe0740d3a","year":2004},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"A. Herschtal and B. Raskutti. Optimising area under the roc curve using gradient descent. In ICML ’04: Proceedings of the twenty-first inter- national conference on Machine learning, page 49, New York,","work_id":"16241d65-2ae4-49a1-a313-d9aebc422590","year":2004},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"T. Hofmann. Latent semantic models for collabo- rative filtering. ACM Trans. Inf. Syst. , 22(1):89– 115, 2004","work_id":"30cb0301-626d-4e11-8402-99f2239c3cf6","year":2004},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Y. Hu, Y. Koren, and C. Volinsky. Collaborative filtering for implicit feedback datasets. In IEEE International Conference on Data Mining (ICDM 2008), pages 263–272, 2008","work_id":"9e666f2d-7918-4a97-89ba-fd0fb71ea901","year":2008}],"snapshot_sha256":"e84876f435cf0296916d908c4a2ced6cd66a1d998113a001750031c130ceb9eb"},"source":{"id":"1205.2618","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-13T22:59:32.218731Z","id":"11bcfaf3-975b-4df5-8f6c-1fe2fb34c349","model_set":{"reader":"grok-4.3"},"one_line_summary":"BPR-Opt is a maximum posterior estimator for personalized ranking from implicit feedback that, when used to train matrix factorization and adaptive kNN, outperforms standard learning techniques on ranking tasks.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Bayesian Personalized Ranking derives an optimization criterion that directly targets ranking from implicit feedback.","strongest_claim":"Our experiments indicate that for the task of personalized ranking our optimization method outperforms the standard learning techniques for MF and kNN.","weakest_assumption":"The assumption that maximizing the BPR-Opt posterior (derived from a Bayesian analysis of pairwise preferences) produces better ranking performance than standard losses when applied to matrix factorization and kNN models."}},"verdict_id":"11bcfaf3-975b-4df5-8f6c-1fe2fb34c349"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:3800ad311e853618ee9c60cfaaa2805267af96b257e12e1c646e6e09b683f7bb","target":"record","created_at":"2026-05-18T03:55:47Z","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":"892b0d523a721ff71fc0c2bb6c9327782bb694b8845a702c7f431065a4578439","cross_cats_sorted":["cs.LG","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2012-05-09T18:25:09Z","title_canon_sha256":"a2ae5ee8235f28fc19ef50476f72eb3c83c25f5b36bb5546e91b38660f177512"},"schema_version":"1.0","source":{"id":"1205.2618","kind":"arxiv","version":1}},"canonical_sha256":"7a11d876d5f4973e13b3d7562cc4c0d024ea0678b23167d1feb3501d273c707a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7a11d876d5f4973e13b3d7562cc4c0d024ea0678b23167d1feb3501d273c707a","first_computed_at":"2026-05-18T03:55:47.886750Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T03:55:47.886750Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4w/Q/bjqxMAFcY5Y0G1sghsyhsRFNMvX6wIZUHVm/PDu+zdSOcX9QqoVIGl0paiQmUILNkZgcFpKg+IvdZ1SBw==","signature_status":"signed_v1","signed_at":"2026-05-18T03:55:47.887260Z","signed_message":"canonical_sha256_bytes"},"source_id":"1205.2618","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3800ad311e853618ee9c60cfaaa2805267af96b257e12e1c646e6e09b683f7bb","sha256:f721c8a1bbcff204f344617174aa71bae5e7ce4aac1a9a864bcec09a0b0d0a3c"],"state_sha256":"9ec4a209787f9ff580e6be92f20d88c58acee97c649f05a2d2e08319158f4cd7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"qgwcCpVxy13nPJO+JvvN67/gYO5kSskwqJS8U48NaFsQVD5FHLQpha7OUeIawTkiiWm7kYlMdS2GoMjT47ECDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-25T16:40:56.842424Z","bundle_sha256":"4ae98994daa95908064f9abc655f2ef4437a731a4fea246f195e6c159bd1c4d5"}}