{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2017:I345H75Q2A5BN6ZUDPV3VWAOUC","short_pith_number":"pith:I345H75Q","canonical_record":{"source":{"id":"1709.00103","kind":"arxiv","version":7},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2017-08-31T23:12:15Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1a5ccd4fc7bdb0c7f10e9cb3ff6460df4ab865e742bccf6b281a0bf23dd02796","abstract_canon_sha256":"634462eb8ea9f932ce45f70ef44f9c4918af5f660aadc44e312a5c09d7767d75"},"schema_version":"1.0"},"canonical_sha256":"46f9d3ffb0d03a16fb341bebbad80ea0a47b4062b97061407c38783053958c4b","source":{"kind":"arxiv","id":"1709.00103","version":7},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1709.00103","created_at":"2026-07-04T22:19:56Z"},{"alias_kind":"arxiv_version","alias_value":"1709.00103v7","created_at":"2026-07-04T22:19:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1709.00103","created_at":"2026-07-04T22:19:56Z"},{"alias_kind":"pith_short_12","alias_value":"I345H75Q2A5B","created_at":"2026-07-04T22:19:56Z"},{"alias_kind":"pith_short_16","alias_value":"I345H75Q2A5BN6ZU","created_at":"2026-07-04T22:19:56Z"},{"alias_kind":"pith_short_8","alias_value":"I345H75Q","created_at":"2026-07-04T22:19:56Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2017:I345H75Q2A5BN6ZUDPV3VWAOUC","target":"record","payload":{"canonical_record":{"source":{"id":"1709.00103","kind":"arxiv","version":7},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2017-08-31T23:12:15Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1a5ccd4fc7bdb0c7f10e9cb3ff6460df4ab865e742bccf6b281a0bf23dd02796","abstract_canon_sha256":"634462eb8ea9f932ce45f70ef44f9c4918af5f660aadc44e312a5c09d7767d75"},"schema_version":"1.0"},"canonical_sha256":"46f9d3ffb0d03a16fb341bebbad80ea0a47b4062b97061407c38783053958c4b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-04T22:19:56.634397Z","signature_b64":"A1h6dUWEZwUK+3Q4qy/lSzm1OlsxkdRwQz2dlpx+Z0i0/enfsSnNibXSuu2P8aZgOYqMks5USjbxO3YhnH1MBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"46f9d3ffb0d03a16fb341bebbad80ea0a47b4062b97061407c38783053958c4b","last_reissued_at":"2026-07-04T22:19:56.633883Z","signature_status":"signed_v1","first_computed_at":"2026-07-04T22:19:56.633883Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1709.00103","source_version":7,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-04T22:19:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"axO8WBVt7gsOijM0sA3unjcu5CyazaDpDwZSUryDZXfiMN88uPBUFr44csWhOSdem3UNkwGWfn1m5IijiFOMCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-13T22:18:15.603290Z"},"content_sha256":"01f2e4e4713ae5015e5c9aae6450d5bf9aebc0bcf1ed721e9d346712bc2d38e1","schema_version":"1.0","event_id":"sha256:01f2e4e4713ae5015e5c9aae6450d5bf9aebc0bcf1ed721e9d346712bc2d38e1"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2017:I345H75Q2A5BN6ZUDPV3VWAOUC","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Seq2SQL: Generating Structured Queries from Natural Language using Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Seq2SQL translates natural language questions into SQL queries by combining structured generation with reinforcement learning rewards from database executions, reaching 59.4 percent execution accuracy.","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Caiming Xiong, Richard Socher, Victor Zhong","submitted_at":"2017-08-31T23:12:15Z","abstract_excerpt":"A significant amount of the world's knowledge is stored in relational databases. However, the ability for users to retrieve facts from a database is limited due to a lack of understanding of query languages such as SQL. We propose Seq2SQL, a deep neural network for translating natural language questions to corresponding SQL queries. Our model leverages the structure of SQL queries to significantly reduce the output space of generated queries. Moreover, we use rewards from in-the-loop query execution over the database to learn a policy to generate unordered parts of the query, which we show are"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"By applying policy-based reinforcement learning with a query execution environment to WikiSQL, our model Seq2SQL outperforms attentional sequence to sequence models, improving execution accuracy from 35.9% to 59.4% and logical form accuracy from 23.4% to 48.3%.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That rewards obtained by executing generated queries on the database provide a sufficiently dense and stable training signal for the policy, especially for the unordered components of SQL.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Seq2SQL uses deep learning plus reinforcement learning to generate SQL from natural language, reaching 59.4% execution accuracy on the new WikiSQL dataset of 80k examples.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Seq2SQL translates natural language questions into SQL queries by combining structured generation with reinforcement learning rewards from database executions, reaching 59.4 percent execution accuracy.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"02f1cab606ebe168e11c252ab744f574a9322c5c490527d61641dffcc1c5421a"},"source":{"id":"1709.00103","kind":"arxiv","version":7},"verdict":{"id":"d180eb3c-e44e-413e-96a7-b3fdb5a9211e","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-13T16:54:59.233529Z","strongest_claim":"By applying policy-based reinforcement learning with a query execution environment to WikiSQL, our model Seq2SQL outperforms attentional sequence to sequence models, improving execution accuracy from 35.9% to 59.4% and logical form accuracy from 23.4% to 48.3%.","one_line_summary":"Seq2SQL uses deep learning plus reinforcement learning to generate SQL from natural language, reaching 59.4% execution accuracy on the new WikiSQL dataset of 80k examples.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That rewards obtained by executing generated queries on the database provide a sufficiently dense and stable training signal for the policy, especially for the unordered components of SQL.","pith_extraction_headline":"Seq2SQL translates natural language questions into SQL queries by combining structured generation with reinforcement learning rewards from database executions, reaching 59.4 percent execution accuracy."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/1709.00103/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":43,"sample":[{"doi":"","year":1995,"title":"I. Androutsopoulos, G.D. Ritchie, and P. Thanisch. Natural language interfaces to databases - an introduction. 1995","work_id":"b33e946d-1f29-43dc-9388-9de4e63852bd","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2011,"title":"Yoav Artzi and Luke S. Zettlemoyer. Bootstrapping semantic parsers from conversations. In EMNLP, 2011","work_id":"8cec51c2-7aed-4594-9bca-c207c75212e7","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2013,"title":"Yoav Artzi and Luke S. Zettlemoyer. Weakly supervised learning of semantic parsers for mapping instructions to actions. TACL, 1: 0 49--62, 2013","work_id":"56b3e86a-36ce-43ce-b295-2d61c2794fe4","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2000,"title":"A new database on the structure and development of the financial sector","work_id":"aa94016f-5f82-45af-9cc7-6f5d64da7964","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2017,"title":"Le, Mohammad Norouzi, and Samy Bengio","work_id":"ee8f2221-c26b-4a37-a4cb-52171d86b086","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":43,"snapshot_sha256":"95a4ae75f9b9ca5ef3905eb4263173fde1cd546e2a69aff7d5ed0bef82e7813e","internal_anchors":1},"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":"d180eb3c-e44e-413e-96a7-b3fdb5a9211e"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-04T22:19:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Eo0wAPOAsdZtKT+3tGWQSkp1EnG+b8c3HFnTzJNvhHyATGVySp+u6aEby2laeSzHUngSES08FAaaKgQzGkzKAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-13T22:18:15.603847Z"},"content_sha256":"595924da241999ff2b74d02f0805c41dc3c4ce47898ec185314b4462226071b5","schema_version":"1.0","event_id":"sha256:595924da241999ff2b74d02f0805c41dc3c4ce47898ec185314b4462226071b5"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/I345H75Q2A5BN6ZUDPV3VWAOUC/bundle.json","state_url":"https://pith.science/pith/I345H75Q2A5BN6ZUDPV3VWAOUC/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/I345H75Q2A5BN6ZUDPV3VWAOUC/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":"2026-07-13T22:18:15Z","links":{"resolver":"https://pith.science/pith/I345H75Q2A5BN6ZUDPV3VWAOUC","bundle":"https://pith.science/pith/I345H75Q2A5BN6ZUDPV3VWAOUC/bundle.json","state":"https://pith.science/pith/I345H75Q2A5BN6ZUDPV3VWAOUC/state.json","well_known_bundle":"https://pith.science/.well-known/pith/I345H75Q2A5BN6ZUDPV3VWAOUC/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2017:I345H75Q2A5BN6ZUDPV3VWAOUC","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":"634462eb8ea9f932ce45f70ef44f9c4918af5f660aadc44e312a5c09d7767d75","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2017-08-31T23:12:15Z","title_canon_sha256":"1a5ccd4fc7bdb0c7f10e9cb3ff6460df4ab865e742bccf6b281a0bf23dd02796"},"schema_version":"1.0","source":{"id":"1709.00103","kind":"arxiv","version":7}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1709.00103","created_at":"2026-07-04T22:19:56Z"},{"alias_kind":"arxiv_version","alias_value":"1709.00103v7","created_at":"2026-07-04T22:19:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1709.00103","created_at":"2026-07-04T22:19:56Z"},{"alias_kind":"pith_short_12","alias_value":"I345H75Q2A5B","created_at":"2026-07-04T22:19:56Z"},{"alias_kind":"pith_short_16","alias_value":"I345H75Q2A5BN6ZU","created_at":"2026-07-04T22:19:56Z"},{"alias_kind":"pith_short_8","alias_value":"I345H75Q","created_at":"2026-07-04T22:19:56Z"}],"graph_snapshots":[{"event_id":"sha256:595924da241999ff2b74d02f0805c41dc3c4ce47898ec185314b4462226071b5","target":"graph","created_at":"2026-07-04T22:19:56Z","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":"By applying policy-based reinforcement learning with a query execution environment to WikiSQL, our model Seq2SQL outperforms attentional sequence to sequence models, improving execution accuracy from 35.9% to 59.4% and logical form accuracy from 23.4% to 48.3%."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That rewards obtained by executing generated queries on the database provide a sufficiently dense and stable training signal for the policy, especially for the unordered components of SQL."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Seq2SQL uses deep learning plus reinforcement learning to generate SQL from natural language, reaching 59.4% execution accuracy on the new WikiSQL dataset of 80k examples."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Seq2SQL translates natural language questions into SQL queries by combining structured generation with reinforcement learning rewards from database executions, reaching 59.4 percent execution accuracy."}],"snapshot_sha256":"02f1cab606ebe168e11c252ab744f574a9322c5c490527d61641dffcc1c5421a"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/1709.00103/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"A significant amount of the world's knowledge is stored in relational databases. However, the ability for users to retrieve facts from a database is limited due to a lack of understanding of query languages such as SQL. We propose Seq2SQL, a deep neural network for translating natural language questions to corresponding SQL queries. Our model leverages the structure of SQL queries to significantly reduce the output space of generated queries. Moreover, we use rewards from in-the-loop query execution over the database to learn a policy to generate unordered parts of the query, which we show are","authors_text":"Caiming Xiong, Richard Socher, Victor Zhong","cross_cats":["cs.AI"],"headline":"Seq2SQL translates natural language questions into SQL queries by combining structured generation with reinforcement learning rewards from database executions, reaching 59.4 percent execution accuracy.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2017-08-31T23:12:15Z","title":"Seq2SQL: Generating Structured Queries from Natural Language using Reinforcement Learning"},"references":{"count":43,"internal_anchors":1,"resolved_work":43,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"I. Androutsopoulos, G.D. Ritchie, and P. Thanisch. Natural language interfaces to databases - an introduction. 1995","work_id":"b33e946d-1f29-43dc-9388-9de4e63852bd","year":1995},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Yoav Artzi and Luke S. Zettlemoyer. Bootstrapping semantic parsers from conversations. In EMNLP, 2011","work_id":"8cec51c2-7aed-4594-9bca-c207c75212e7","year":2011},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Yoav Artzi and Luke S. Zettlemoyer. Weakly supervised learning of semantic parsers for mapping instructions to actions. TACL, 1: 0 49--62, 2013","work_id":"56b3e86a-36ce-43ce-b295-2d61c2794fe4","year":2013},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"A new database on the structure and development of the financial sector","work_id":"aa94016f-5f82-45af-9cc7-6f5d64da7964","year":2000},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Le, Mohammad Norouzi, and Samy Bengio","work_id":"ee8f2221-c26b-4a37-a4cb-52171d86b086","year":2017}],"snapshot_sha256":"95a4ae75f9b9ca5ef3905eb4263173fde1cd546e2a69aff7d5ed0bef82e7813e"},"source":{"id":"1709.00103","kind":"arxiv","version":7},"verdict":{"created_at":"2026-05-13T16:54:59.233529Z","id":"d180eb3c-e44e-413e-96a7-b3fdb5a9211e","model_set":{"reader":"grok-4.3"},"one_line_summary":"Seq2SQL uses deep learning plus reinforcement learning to generate SQL from natural language, reaching 59.4% execution accuracy on the new WikiSQL dataset of 80k examples.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Seq2SQL translates natural language questions into SQL queries by combining structured generation with reinforcement learning rewards from database executions, reaching 59.4 percent execution accuracy.","strongest_claim":"By applying policy-based reinforcement learning with a query execution environment to WikiSQL, our model Seq2SQL outperforms attentional sequence to sequence models, improving execution accuracy from 35.9% to 59.4% and logical form accuracy from 23.4% to 48.3%.","weakest_assumption":"That rewards obtained by executing generated queries on the database provide a sufficiently dense and stable training signal for the policy, especially for the unordered components of SQL."}},"verdict_id":"d180eb3c-e44e-413e-96a7-b3fdb5a9211e"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:01f2e4e4713ae5015e5c9aae6450d5bf9aebc0bcf1ed721e9d346712bc2d38e1","target":"record","created_at":"2026-07-04T22:19:56Z","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":"634462eb8ea9f932ce45f70ef44f9c4918af5f660aadc44e312a5c09d7767d75","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2017-08-31T23:12:15Z","title_canon_sha256":"1a5ccd4fc7bdb0c7f10e9cb3ff6460df4ab865e742bccf6b281a0bf23dd02796"},"schema_version":"1.0","source":{"id":"1709.00103","kind":"arxiv","version":7}},"canonical_sha256":"46f9d3ffb0d03a16fb341bebbad80ea0a47b4062b97061407c38783053958c4b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"46f9d3ffb0d03a16fb341bebbad80ea0a47b4062b97061407c38783053958c4b","first_computed_at":"2026-07-04T22:19:56.633883Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-04T22:19:56.633883Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"A1h6dUWEZwUK+3Q4qy/lSzm1OlsxkdRwQz2dlpx+Z0i0/enfsSnNibXSuu2P8aZgOYqMks5USjbxO3YhnH1MBA==","signature_status":"signed_v1","signed_at":"2026-07-04T22:19:56.634397Z","signed_message":"canonical_sha256_bytes"},"source_id":"1709.00103","source_kind":"arxiv","source_version":7}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:01f2e4e4713ae5015e5c9aae6450d5bf9aebc0bcf1ed721e9d346712bc2d38e1","sha256:595924da241999ff2b74d02f0805c41dc3c4ce47898ec185314b4462226071b5"],"state_sha256":"59429c5606770b124614a29012ebd04fea2e21e0cbf5ac7ab74b9b4e7e637beb"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FTtYhHBgYVBaI99YViKPJAhO2QOqadLFv93OYcz1fq+ttlAv6x8LTTdIaHbyg7fnIHr0l56H543+R/EM+Wi4CA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-13T22:18:15.606716Z","bundle_sha256":"8649441b901abaabca47e2853800a19dbf7c294e173939c2498839d4c023c52f"}}