{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:JXLOLHJTZL3YVPFKZ4N4FI6BNS","short_pith_number":"pith:JXLOLHJT","canonical_record":{"source":{"id":"2602.13155","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-02-13T18:08:23Z","cross_cats_sorted":["cs.DS","cs.NE","stat.ML"],"title_canon_sha256":"642040a87493a33ba1557d66e53c2e71d3f799821b2fad64c60cd5ae24c21a1a","abstract_canon_sha256":"5ed359c114420460314653bb65424698df15b12343c7a3684cb1b5f8a74c7bd5"},"schema_version":"1.0"},"canonical_sha256":"4dd6e59d33caf78abcaacf1bc2a3c16c9162027e4f7714a6fe4b8a0aa12c8cf6","source":{"kind":"arxiv","id":"2602.13155","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.13155","created_at":"2026-05-18T02:45:05Z"},{"alias_kind":"arxiv_version","alias_value":"2602.13155v2","created_at":"2026-05-18T02:45:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.13155","created_at":"2026-05-18T02:45:05Z"},{"alias_kind":"pith_short_12","alias_value":"JXLOLHJTZL3Y","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"JXLOLHJTZL3YVPFK","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"JXLOLHJT","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:JXLOLHJTZL3YVPFKZ4N4FI6BNS","target":"record","payload":{"canonical_record":{"source":{"id":"2602.13155","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-02-13T18:08:23Z","cross_cats_sorted":["cs.DS","cs.NE","stat.ML"],"title_canon_sha256":"642040a87493a33ba1557d66e53c2e71d3f799821b2fad64c60cd5ae24c21a1a","abstract_canon_sha256":"5ed359c114420460314653bb65424698df15b12343c7a3684cb1b5f8a74c7bd5"},"schema_version":"1.0"},"canonical_sha256":"4dd6e59d33caf78abcaacf1bc2a3c16c9162027e4f7714a6fe4b8a0aa12c8cf6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:45:05.257763Z","signature_b64":"Wwt1x0GpCdUWMTrJetLBNRtyeZNtmgSZc3G+fFnezaTUb9sDpuu81tp0vlfgU8HJHGPeXL6jH5bx3cA6HENnCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4dd6e59d33caf78abcaacf1bc2a3c16c9162027e4f7714a6fe4b8a0aa12c8cf6","last_reissued_at":"2026-05-18T02:45:05.257246Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:45:05.257246Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2602.13155","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:45:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Ryu3YLh3oIznqcZRedaT12QUEOAUuQMWxzCOSFxvK+Ki2ZwnJO2uvnvVpA5wsuWzZD5AAXTua8On53dEdRGQCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-08T04:28:10.060271Z"},"content_sha256":"ddddc4fe8dfebbbc0a4d23cd8f945282bccdec3ee67554cf6b64215eac0e9d6a","schema_version":"1.0","event_id":"sha256:ddddc4fe8dfebbbc0a4d23cd8f945282bccdec3ee67554cf6b64215eac0e9d6a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:JXLOLHJTZL3YVPFKZ4N4FI6BNS","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Learning to Approximate Uniform Facility Location via Graph Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A message-passing neural network embeds approximation principles to solve uniform facility location with guarantees and better empirical performance.","cross_cats":["cs.DS","cs.NE","stat.ML"],"primary_cat":"cs.LG","authors_text":"Chendi Qian, Christian Sohler, Christopher Morris, Stefanie Jegelka","submitted_at":"2026-02-13T18:08:23Z","abstract_excerpt":"Neural networks, particularly message-passing neural networks (MPNNs), are increasingly used as heuristics for hard combinatorial optimization problems. Yet many learning-based methods rely on supervision, reinforcement learning, or gradient estimators, causing high computational cost, unstable training, or limited guarantees. Classical approximation algorithms provide worst-case guarantees but are non-differentiable and cannot adapt to structure in natural input distributions. We study this tradeoff through Uniform Facility Location (UniFL), a problem with applications in clustering, summariz"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We propose a fully differentiable MPNN that incorporates approximation-algorithmic principles without solver supervision or discrete relaxations. The model has provable approximation guarantees and empirically improves on standard approximation algorithms, narrowing the gap to integer linear programming.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That embedding approximation-algorithmic principles into an MPNN preserves the provable guarantees while allowing the network to improve empirically on the target distribution without any supervision or discrete relaxations.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A differentiable MPNN approximates uniform facility location with provable guarantees and outperforms standard approximation algorithms while closing the gap to exact ILP solutions.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A message-passing neural network embeds approximation principles to solve uniform facility location with guarantees and better empirical performance.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"fb372603980e69ad1195cea5207874535a22454f8e96f447715ebbf5ac6bd7de"},"source":{"id":"2602.13155","kind":"arxiv","version":2},"verdict":{"id":"4264aca0-2497-46e1-b03a-50e998ee3494","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T22:12:03.613354Z","strongest_claim":"We propose a fully differentiable MPNN that incorporates approximation-algorithmic principles without solver supervision or discrete relaxations. The model has provable approximation guarantees and empirically improves on standard approximation algorithms, narrowing the gap to integer linear programming.","one_line_summary":"A differentiable MPNN approximates uniform facility location with provable guarantees and outperforms standard approximation algorithms while closing the gap to exact ILP solutions.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That embedding approximation-algorithmic principles into an MPNN preserves the provable guarantees while allowing the network to improve empirically on the target distribution without any supervision or discrete relaxations.","pith_extraction_headline":"A message-passing neural network embeds approximation principles to solve uniform facility location with guarantees and better empirical performance."},"references":{"count":15,"sample":[{"doi":"","year":null,"title":"3, 4, 5 R. B. Bairi, R. Iyer, G. Ramakrishnan, and J. Bilmes. Summarization of multi-document topic hierarchies using submodular mixtures. InAssociation of Computational Linguists (ACL),","work_id":"26614e6a-dd33-4197-b474-99ffb29cf50c","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":1997,"title":"Neural Combinatorial Optimization with Reinforcement Learning","work_id":"86c3352e-7964-488c-9a6a-22803e1cd602","ref_index":2,"cited_arxiv_id":"1611.09940","is_internal_anchor":true},{"doi":"10.1145/375827.375845","year":2020,"title":"Simpler analyses of local search algorithms for facility loca- tion","work_id":"ccf67ae4-2c5d-4f8e-8604-910844819020","ref_index":3,"cited_arxiv_id":"0809.2554","is_internal_anchor":true},{"doi":"","year":2022,"title":"2 S. H. Jiang, E. Liu, Y. Lyu, Z. G. Tang, and Y. Zhang. Online facility location with predictions. InICLR, 2022. 23 N. Karalias and A. Loukas. Erdos goes neural: an unsupervised learning framework fo","work_id":"41a6ea61-2db5-49f7-a9d8-ad0f5510a023","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1137/s0097539703435716","year":2025,"title":"23 E. Levin, Y. Ma, M. Díaz, and S. Villar. On transferring transferability: Towards a theory for size generalization.arXiv preprint arXiv:2505.23599, 2025. 23 17 H. Liang, H. S. d. O. Borde, B. Sripa","work_id":"3741303e-3a6c-4999-ac5a-d056278e5e1d","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":15,"snapshot_sha256":"5f6733158f3b5c8ff35645ae1c7450608bb2683c5747dc42642b22d3e8b1e18c","internal_anchors":3},"formal_canon":{"evidence_count":2,"snapshot_sha256":"917dc1750d6c3bf82e3f9589c7a90f7dd46fc9cc46297c0edd5a67ae3a2065b3"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"4264aca0-2497-46e1-b03a-50e998ee3494"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:45:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"qzmriYUFBy88JXbV5hbEGU3nREMGJ0UX9JBnkTHPMxPC5HZvveyLUGLUuYq2WMaZyxgLXe4Mz4Apsj+IVyWZAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-08T04:28:10.061303Z"},"content_sha256":"9765652101c54865ed5b9b02c27520a2d8081082be66162bc6a245d9fb71758f","schema_version":"1.0","event_id":"sha256:9765652101c54865ed5b9b02c27520a2d8081082be66162bc6a245d9fb71758f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JXLOLHJTZL3YVPFKZ4N4FI6BNS/bundle.json","state_url":"https://pith.science/pith/JXLOLHJTZL3YVPFKZ4N4FI6BNS/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JXLOLHJTZL3YVPFKZ4N4FI6BNS/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":"2026-06-08T04:28:10Z","links":{"resolver":"https://pith.science/pith/JXLOLHJTZL3YVPFKZ4N4FI6BNS","bundle":"https://pith.science/pith/JXLOLHJTZL3YVPFKZ4N4FI6BNS/bundle.json","state":"https://pith.science/pith/JXLOLHJTZL3YVPFKZ4N4FI6BNS/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JXLOLHJTZL3YVPFKZ4N4FI6BNS/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:JXLOLHJTZL3YVPFKZ4N4FI6BNS","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":"5ed359c114420460314653bb65424698df15b12343c7a3684cb1b5f8a74c7bd5","cross_cats_sorted":["cs.DS","cs.NE","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-02-13T18:08:23Z","title_canon_sha256":"642040a87493a33ba1557d66e53c2e71d3f799821b2fad64c60cd5ae24c21a1a"},"schema_version":"1.0","source":{"id":"2602.13155","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.13155","created_at":"2026-05-18T02:45:05Z"},{"alias_kind":"arxiv_version","alias_value":"2602.13155v2","created_at":"2026-05-18T02:45:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.13155","created_at":"2026-05-18T02:45:05Z"},{"alias_kind":"pith_short_12","alias_value":"JXLOLHJTZL3Y","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"JXLOLHJTZL3YVPFK","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"JXLOLHJT","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:9765652101c54865ed5b9b02c27520a2d8081082be66162bc6a245d9fb71758f","target":"graph","created_at":"2026-05-18T02:45:05Z","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":"We propose a fully differentiable MPNN that incorporates approximation-algorithmic principles without solver supervision or discrete relaxations. The model has provable approximation guarantees and empirically improves on standard approximation algorithms, narrowing the gap to integer linear programming."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That embedding approximation-algorithmic principles into an MPNN preserves the provable guarantees while allowing the network to improve empirically on the target distribution without any supervision or discrete relaxations."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"A differentiable MPNN approximates uniform facility location with provable guarantees and outperforms standard approximation algorithms while closing the gap to exact ILP solutions."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"A message-passing neural network embeds approximation principles to solve uniform facility location with guarantees and better empirical performance."}],"snapshot_sha256":"fb372603980e69ad1195cea5207874535a22454f8e96f447715ebbf5ac6bd7de"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"917dc1750d6c3bf82e3f9589c7a90f7dd46fc9cc46297c0edd5a67ae3a2065b3"},"paper":{"abstract_excerpt":"Neural networks, particularly message-passing neural networks (MPNNs), are increasingly used as heuristics for hard combinatorial optimization problems. Yet many learning-based methods rely on supervision, reinforcement learning, or gradient estimators, causing high computational cost, unstable training, or limited guarantees. Classical approximation algorithms provide worst-case guarantees but are non-differentiable and cannot adapt to structure in natural input distributions. We study this tradeoff through Uniform Facility Location (UniFL), a problem with applications in clustering, summariz","authors_text":"Chendi Qian, Christian Sohler, Christopher Morris, Stefanie Jegelka","cross_cats":["cs.DS","cs.NE","stat.ML"],"headline":"A message-passing neural network embeds approximation principles to solve uniform facility location with guarantees and better empirical performance.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-02-13T18:08:23Z","title":"Learning to Approximate Uniform Facility Location via Graph Neural Networks"},"references":{"count":15,"internal_anchors":3,"resolved_work":15,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"3, 4, 5 R. B. Bairi, R. Iyer, G. Ramakrishnan, and J. Bilmes. Summarization of multi-document topic hierarchies using submodular mixtures. InAssociation of Computational Linguists (ACL),","work_id":"26614e6a-dd33-4197-b474-99ffb29cf50c","year":null},{"cited_arxiv_id":"1611.09940","doi":"","is_internal_anchor":true,"ref_index":2,"title":"Neural Combinatorial Optimization with Reinforcement Learning","work_id":"86c3352e-7964-488c-9a6a-22803e1cd602","year":1997},{"cited_arxiv_id":"0809.2554","doi":"10.1145/375827.375845","is_internal_anchor":true,"ref_index":3,"title":"Simpler analyses of local search algorithms for facility loca- tion","work_id":"ccf67ae4-2c5d-4f8e-8604-910844819020","year":2020},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"2 S. H. Jiang, E. Liu, Y. Lyu, Z. G. Tang, and Y. Zhang. Online facility location with predictions. InICLR, 2022. 23 N. Karalias and A. Loukas. Erdos goes neural: an unsupervised learning framework fo","work_id":"41a6ea61-2db5-49f7-a9d8-ad0f5510a023","year":2022},{"cited_arxiv_id":"","doi":"10.1137/s0097539703435716","is_internal_anchor":false,"ref_index":5,"title":"23 E. Levin, Y. Ma, M. Díaz, and S. Villar. On transferring transferability: Towards a theory for size generalization.arXiv preprint arXiv:2505.23599, 2025. 23 17 H. Liang, H. S. d. O. Borde, B. Sripa","work_id":"3741303e-3a6c-4999-ac5a-d056278e5e1d","year":2025}],"snapshot_sha256":"5f6733158f3b5c8ff35645ae1c7450608bb2683c5747dc42642b22d3e8b1e18c"},"source":{"id":"2602.13155","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-15T22:12:03.613354Z","id":"4264aca0-2497-46e1-b03a-50e998ee3494","model_set":{"reader":"grok-4.3"},"one_line_summary":"A differentiable MPNN approximates uniform facility location with provable guarantees and outperforms standard approximation algorithms while closing the gap to exact ILP solutions.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"A message-passing neural network embeds approximation principles to solve uniform facility location with guarantees and better empirical performance.","strongest_claim":"We propose a fully differentiable MPNN that incorporates approximation-algorithmic principles without solver supervision or discrete relaxations. The model has provable approximation guarantees and empirically improves on standard approximation algorithms, narrowing the gap to integer linear programming.","weakest_assumption":"That embedding approximation-algorithmic principles into an MPNN preserves the provable guarantees while allowing the network to improve empirically on the target distribution without any supervision or discrete relaxations."}},"verdict_id":"4264aca0-2497-46e1-b03a-50e998ee3494"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:ddddc4fe8dfebbbc0a4d23cd8f945282bccdec3ee67554cf6b64215eac0e9d6a","target":"record","created_at":"2026-05-18T02:45:05Z","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":"5ed359c114420460314653bb65424698df15b12343c7a3684cb1b5f8a74c7bd5","cross_cats_sorted":["cs.DS","cs.NE","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-02-13T18:08:23Z","title_canon_sha256":"642040a87493a33ba1557d66e53c2e71d3f799821b2fad64c60cd5ae24c21a1a"},"schema_version":"1.0","source":{"id":"2602.13155","kind":"arxiv","version":2}},"canonical_sha256":"4dd6e59d33caf78abcaacf1bc2a3c16c9162027e4f7714a6fe4b8a0aa12c8cf6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4dd6e59d33caf78abcaacf1bc2a3c16c9162027e4f7714a6fe4b8a0aa12c8cf6","first_computed_at":"2026-05-18T02:45:05.257246Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T02:45:05.257246Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Wwt1x0GpCdUWMTrJetLBNRtyeZNtmgSZc3G+fFnezaTUb9sDpuu81tp0vlfgU8HJHGPeXL6jH5bx3cA6HENnCg==","signature_status":"signed_v1","signed_at":"2026-05-18T02:45:05.257763Z","signed_message":"canonical_sha256_bytes"},"source_id":"2602.13155","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ddddc4fe8dfebbbc0a4d23cd8f945282bccdec3ee67554cf6b64215eac0e9d6a","sha256:9765652101c54865ed5b9b02c27520a2d8081082be66162bc6a245d9fb71758f"],"state_sha256":"26e9e2289f1108359ff65b0e9f8a8e98ecfda4cc0972220b353eab6e97cb1d15"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ycEC5lb7n1TS+V7cGHtCStMfCqO5nKJYu+14yd12YOuHmximBgVngmUjpqwTy5+JQXa+ctGfQQbR9RexH+pQDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-08T04:28:10.066147Z","bundle_sha256":"d4749c07173c39f7627150a468f3be1145de1e463088d4c41900c7ed79e3bf7c"}}