{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:IR65NTO76M2AKSLFKIDN7GDDMR","short_pith_number":"pith:IR65NTO7","canonical_record":{"source":{"id":"2605.13181","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T08:41:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ba7c6a38f3d0f8323ede6c3dba9486f91367342a7dce476b97d24c1866563a6b","abstract_canon_sha256":"e794eef84cdbeb9e018d1455fc114f549e55f21a109124d39cc385cab878b6d2"},"schema_version":"1.0"},"canonical_sha256":"447dd6cddff3340549655206df986364551a7b1f4f778c4c234d55cd7cf3dd17","source":{"kind":"arxiv","id":"2605.13181","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13181","created_at":"2026-05-18T03:08:56Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13181v1","created_at":"2026-05-18T03:08:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13181","created_at":"2026-05-18T03:08:56Z"},{"alias_kind":"pith_short_12","alias_value":"IR65NTO76M2A","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"IR65NTO76M2AKSLF","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"IR65NTO7","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:IR65NTO76M2AKSLFKIDN7GDDMR","target":"record","payload":{"canonical_record":{"source":{"id":"2605.13181","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T08:41:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ba7c6a38f3d0f8323ede6c3dba9486f91367342a7dce476b97d24c1866563a6b","abstract_canon_sha256":"e794eef84cdbeb9e018d1455fc114f549e55f21a109124d39cc385cab878b6d2"},"schema_version":"1.0"},"canonical_sha256":"447dd6cddff3340549655206df986364551a7b1f4f778c4c234d55cd7cf3dd17","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T03:08:56.415010Z","signature_b64":"d2CMPRi1JKp79hZIlsaNZdEU9HlLzKGW/t8WcehRbj/qkS3APfzC2VKHQhfx1JgRdc5EPWjs1WV//f76EhFdDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"447dd6cddff3340549655206df986364551a7b1f4f778c4c234d55cd7cf3dd17","last_reissued_at":"2026-05-18T03:08:56.414170Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T03:08:56.414170Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.13181","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":"2026-05-18T03:08:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PY2J1fWg5GFVysvfayT4u347t6WDrrvl1w9dKtggJldeLpLgL+zOdLHvTGS0EUYCrHfqplTW73lvbvQlB8BOCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T13:43:02.860036Z"},"content_sha256":"9b60f6d68fd822274e7e8a0eec3096f2d4b346ab566198c85971b232d966fb50","schema_version":"1.0","event_id":"sha256:9b60f6d68fd822274e7e8a0eec3096f2d4b346ab566198c85971b232d966fb50"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:IR65NTO76M2AKSLFKIDN7GDDMR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Stable Attention Response for Reliable Precipitation Nowcasting","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Cross-sample instability in attention-response energy drives unreliable precipitation nowcasts, which HARECast corrects via group-wise regularization on head-wise energy.","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Allen Benter, Kun Hu, Patrick Filippi, Penghui Wen, Sen Zhang, Thomas Bishop, Xiaogang Zhu, Zexin Hu, Zhiyong Wang","submitted_at":"2026-05-13T08:41:39Z","abstract_excerpt":"Precipitation nowcasting remains challenging due to the highly localized, rapidly evolving, and heterogeneous nature of atmospheric dynamics. Although recent methods increasingly adopt attention-based architectures in both unimodal and multimodal settings, they mainly emphasize stronger representation learning and prediction capacity, while paying less attention to the stability of attention responses across samples. In this work, we show that cross-sample instability of attention-response energy is an important and previously underexplored source of forecasting unreliability. Empirically, ina"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Cross-sample instability of attention-response energy is an important and previously underexplored source of forecasting unreliability; HARECast achieves state-of-the-art performance on SEVIR and MeteoNet by explicitly modeling and stabilizing head-wise attention-response energy through group-wise regularization.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That reducing cross-sample attention energy variance via the proposed regularization will not degrade the model's ability to learn useful representations for accurate precipitation prediction, and that the observed association between energy variance and forecast error is causal rather than correlational.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"HARECast stabilizes cross-sample variance in attention-response energy via group-wise regularization to reduce prediction errors in precipitation nowcasting.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Cross-sample instability in attention-response energy drives unreliable precipitation nowcasts, which HARECast corrects via group-wise regularization on head-wise energy.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"bf072b05f2fcd2bda7961f7d9edf591288b332653c37737bb5a022b2750b47af"},"source":{"id":"2605.13181","kind":"arxiv","version":1},"verdict":{"id":"3cf8bc99-31d9-4b06-8ae4-8d41d5d01971","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T19:16:31.151356Z","strongest_claim":"Cross-sample instability of attention-response energy is an important and previously underexplored source of forecasting unreliability; HARECast achieves state-of-the-art performance on SEVIR and MeteoNet by explicitly modeling and stabilizing head-wise attention-response energy through group-wise regularization.","one_line_summary":"HARECast stabilizes cross-sample variance in attention-response energy via group-wise regularization to reduce prediction errors in precipitation nowcasting.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That reducing cross-sample attention energy variance via the proposed regularization will not degrade the model's ability to learn useful representations for accurate precipitation prediction, and that the observed association between energy variance and forecast error is causal rather than correlational.","pith_extraction_headline":"Cross-sample instability in attention-response energy drives unreliable precipitation nowcasts, which HARECast corrects via group-wise regularization on head-wise energy."},"references":{"count":36,"sample":[{"doi":"","year":2023,"title":"Oussama Boussif, Ghait Boukachab, Dan Assouline, Stefano Massaroli, Tianle Yuan, Loubna Benabbou, and Yoshua Bengio. 2023. Improving day-ahead solar ir- radiance time series forecasting by leveraging ","work_id":"fb9659a3-a03e-4d23-b50c-5f10622ae009","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Zheng Chang, Xinfeng Zhang, Shanshe Wang, Siwei Ma, and Wen Gao. 2022. STRPM: A spatiotemporal residual predictive model for high-resolution video prediction. InConference on Computer Vision and Patte","work_id":"52e38e04-0d0c-4d09-8f1f-db95cd960a88","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"Zheng Chang, Xinfeng Zhang, Shanshe Wang, Siwei Ma, Yan Ye, Xiang Xinguang, and Wen Gao. 2021. MAU: A motion-aware unit for video prediction and beyond. Advances in Neural Information Processing Syste","work_id":"e3548361-4866-4c14-97f2-26d0105b4b3c","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Yeji Choi, Keumgang Cha, Minyoung Back, Hyunguk Choi, and Taegyun Jeon","work_id":"273424d0-315f-4476-8c75-a4688d200649","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"InIEEE International Geoscience and Remote Sensing Symposium","work_id":"48991d52-d5d8-4d8f-b435-51eaa3182b35","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":36,"snapshot_sha256":"bdd63d09cf29348d0ef48c3175479bbff90ba8c6e6bfec8cb12500327bd25dc6","internal_anchors":2},"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":"3cf8bc99-31d9-4b06-8ae4-8d41d5d01971"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T03:08:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"AOEpSh4qU/fQgAfE0eVVhv4jqFIiefv9j39jrK8pCWvOndvDZNqkoS3wqLBnLDwbHIhtTksrXUwz/VgmXdHqBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T13:43:02.861158Z"},"content_sha256":"d04c8ab20dd846ea51e8401142b8e91724f2e5cfc120cff925f702bcc4d293fd","schema_version":"1.0","event_id":"sha256:d04c8ab20dd846ea51e8401142b8e91724f2e5cfc120cff925f702bcc4d293fd"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/IR65NTO76M2AKSLFKIDN7GDDMR/bundle.json","state_url":"https://pith.science/pith/IR65NTO76M2AKSLFKIDN7GDDMR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/IR65NTO76M2AKSLFKIDN7GDDMR/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":"2026-05-25T13:43:02Z","links":{"resolver":"https://pith.science/pith/IR65NTO76M2AKSLFKIDN7GDDMR","bundle":"https://pith.science/pith/IR65NTO76M2AKSLFKIDN7GDDMR/bundle.json","state":"https://pith.science/pith/IR65NTO76M2AKSLFKIDN7GDDMR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/IR65NTO76M2AKSLFKIDN7GDDMR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:IR65NTO76M2AKSLFKIDN7GDDMR","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":"e794eef84cdbeb9e018d1455fc114f549e55f21a109124d39cc385cab878b6d2","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T08:41:39Z","title_canon_sha256":"ba7c6a38f3d0f8323ede6c3dba9486f91367342a7dce476b97d24c1866563a6b"},"schema_version":"1.0","source":{"id":"2605.13181","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13181","created_at":"2026-05-18T03:08:56Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13181v1","created_at":"2026-05-18T03:08:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13181","created_at":"2026-05-18T03:08:56Z"},{"alias_kind":"pith_short_12","alias_value":"IR65NTO76M2A","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"IR65NTO76M2AKSLF","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"IR65NTO7","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:d04c8ab20dd846ea51e8401142b8e91724f2e5cfc120cff925f702bcc4d293fd","target":"graph","created_at":"2026-05-18T03:08:56Z","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":"Cross-sample instability of attention-response energy is an important and previously underexplored source of forecasting unreliability; HARECast achieves state-of-the-art performance on SEVIR and MeteoNet by explicitly modeling and stabilizing head-wise attention-response energy through group-wise regularization."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That reducing cross-sample attention energy variance via the proposed regularization will not degrade the model's ability to learn useful representations for accurate precipitation prediction, and that the observed association between energy variance and forecast error is causal rather than correlational."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"HARECast stabilizes cross-sample variance in attention-response energy via group-wise regularization to reduce prediction errors in precipitation nowcasting."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Cross-sample instability in attention-response energy drives unreliable precipitation nowcasts, which HARECast corrects via group-wise regularization on head-wise energy."}],"snapshot_sha256":"bf072b05f2fcd2bda7961f7d9edf591288b332653c37737bb5a022b2750b47af"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"Precipitation nowcasting remains challenging due to the highly localized, rapidly evolving, and heterogeneous nature of atmospheric dynamics. Although recent methods increasingly adopt attention-based architectures in both unimodal and multimodal settings, they mainly emphasize stronger representation learning and prediction capacity, while paying less attention to the stability of attention responses across samples. In this work, we show that cross-sample instability of attention-response energy is an important and previously underexplored source of forecasting unreliability. Empirically, ina","authors_text":"Allen Benter, Kun Hu, Patrick Filippi, Penghui Wen, Sen Zhang, Thomas Bishop, Xiaogang Zhu, Zexin Hu, Zhiyong Wang","cross_cats":["cs.AI"],"headline":"Cross-sample instability in attention-response energy drives unreliable precipitation nowcasts, which HARECast corrects via group-wise regularization on head-wise energy.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T08:41:39Z","title":"Stable Attention Response for Reliable Precipitation Nowcasting"},"references":{"count":36,"internal_anchors":2,"resolved_work":36,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Oussama Boussif, Ghait Boukachab, Dan Assouline, Stefano Massaroli, Tianle Yuan, Loubna Benabbou, and Yoshua Bengio. 2023. Improving day-ahead solar ir- radiance time series forecasting by leveraging ","work_id":"fb9659a3-a03e-4d23-b50c-5f10622ae009","year":2023},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Zheng Chang, Xinfeng Zhang, Shanshe Wang, Siwei Ma, and Wen Gao. 2022. STRPM: A spatiotemporal residual predictive model for high-resolution video prediction. InConference on Computer Vision and Patte","work_id":"52e38e04-0d0c-4d09-8f1f-db95cd960a88","year":2022},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Zheng Chang, Xinfeng Zhang, Shanshe Wang, Siwei Ma, Yan Ye, Xiang Xinguang, and Wen Gao. 2021. MAU: A motion-aware unit for video prediction and beyond. Advances in Neural Information Processing Syste","work_id":"e3548361-4866-4c14-97f2-26d0105b4b3c","year":2021},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Yeji Choi, Keumgang Cha, Minyoung Back, Hyunguk Choi, and Taegyun Jeon","work_id":"273424d0-315f-4476-8c75-a4688d200649","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"InIEEE International Geoscience and Remote Sensing Symposium","work_id":"48991d52-d5d8-4d8f-b435-51eaa3182b35","year":null}],"snapshot_sha256":"bdd63d09cf29348d0ef48c3175479bbff90ba8c6e6bfec8cb12500327bd25dc6"},"source":{"id":"2605.13181","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T19:16:31.151356Z","id":"3cf8bc99-31d9-4b06-8ae4-8d41d5d01971","model_set":{"reader":"grok-4.3"},"one_line_summary":"HARECast stabilizes cross-sample variance in attention-response energy via group-wise regularization to reduce prediction errors in precipitation nowcasting.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Cross-sample instability in attention-response energy drives unreliable precipitation nowcasts, which HARECast corrects via group-wise regularization on head-wise energy.","strongest_claim":"Cross-sample instability of attention-response energy is an important and previously underexplored source of forecasting unreliability; HARECast achieves state-of-the-art performance on SEVIR and MeteoNet by explicitly modeling and stabilizing head-wise attention-response energy through group-wise regularization.","weakest_assumption":"That reducing cross-sample attention energy variance via the proposed regularization will not degrade the model's ability to learn useful representations for accurate precipitation prediction, and that the observed association between energy variance and forecast error is causal rather than correlational."}},"verdict_id":"3cf8bc99-31d9-4b06-8ae4-8d41d5d01971"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:9b60f6d68fd822274e7e8a0eec3096f2d4b346ab566198c85971b232d966fb50","target":"record","created_at":"2026-05-18T03:08:56Z","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":"e794eef84cdbeb9e018d1455fc114f549e55f21a109124d39cc385cab878b6d2","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T08:41:39Z","title_canon_sha256":"ba7c6a38f3d0f8323ede6c3dba9486f91367342a7dce476b97d24c1866563a6b"},"schema_version":"1.0","source":{"id":"2605.13181","kind":"arxiv","version":1}},"canonical_sha256":"447dd6cddff3340549655206df986364551a7b1f4f778c4c234d55cd7cf3dd17","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"447dd6cddff3340549655206df986364551a7b1f4f778c4c234d55cd7cf3dd17","first_computed_at":"2026-05-18T03:08:56.414170Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T03:08:56.414170Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"d2CMPRi1JKp79hZIlsaNZdEU9HlLzKGW/t8WcehRbj/qkS3APfzC2VKHQhfx1JgRdc5EPWjs1WV//f76EhFdDg==","signature_status":"signed_v1","signed_at":"2026-05-18T03:08:56.415010Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.13181","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9b60f6d68fd822274e7e8a0eec3096f2d4b346ab566198c85971b232d966fb50","sha256:d04c8ab20dd846ea51e8401142b8e91724f2e5cfc120cff925f702bcc4d293fd"],"state_sha256":"a2f9db86fe1f56cbd6cf9be855bfe6e0fd8989893b7aebd3adf4625e97076f03"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"y6mfuWFc8YBXbA8HI9P7Qr/Xs9v8UasJvO0TFYCBqsYWQGEsr1pK9iZi4KUptjTA528Z9DWABNNkoxz4S6/iAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-25T13:43:02.865806Z","bundle_sha256":"a21fe910b391cd7daebb5a098d383372651f472fe4207035d28c4f0ea1408231"}}