{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:6M6GD7VD5G3YLJT6MSXERFVDKI","short_pith_number":"pith:6M6GD7VD","canonical_record":{"source":{"id":"2512.01242","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-12-01T03:38:44Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"1ceea254e081438fe6352f9c9baff7939cc5b472ceb099fd6d34850737767a8c","abstract_canon_sha256":"eb931f34e2bbcfa6c060e198ba433d82a3a4f392d989f11a1a608b3d34f17d16"},"schema_version":"1.0"},"canonical_sha256":"f33c61fea3e9b785a67e64ae4896a3521a79c5e4b1e62f431599d2212833c4b9","source":{"kind":"arxiv","id":"2512.01242","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2512.01242","created_at":"2026-05-18T03:09:32Z"},{"alias_kind":"arxiv_version","alias_value":"2512.01242v3","created_at":"2026-05-18T03:09:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.01242","created_at":"2026-05-18T03:09:32Z"},{"alias_kind":"pith_short_12","alias_value":"6M6GD7VD5G3Y","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"6M6GD7VD5G3YLJT6","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"6M6GD7VD","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:6M6GD7VD5G3YLJT6MSXERFVDKI","target":"record","payload":{"canonical_record":{"source":{"id":"2512.01242","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-12-01T03:38:44Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"1ceea254e081438fe6352f9c9baff7939cc5b472ceb099fd6d34850737767a8c","abstract_canon_sha256":"eb931f34e2bbcfa6c060e198ba433d82a3a4f392d989f11a1a608b3d34f17d16"},"schema_version":"1.0"},"canonical_sha256":"f33c61fea3e9b785a67e64ae4896a3521a79c5e4b1e62f431599d2212833c4b9","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T03:09:32.924798Z","signature_b64":"aMg31aBYCXfrK+uNGnnQXuL4IWdTh7tIPPAECnLpnmez2MKLF/N8/EzuzcGjy4+oPwIQJVgKyU2UkZNwnptFCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f33c61fea3e9b785a67e64ae4896a3521a79c5e4b1e62f431599d2212833c4b9","last_reissued_at":"2026-05-18T03:09:32.924055Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T03:09:32.924055Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2512.01242","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T03:09:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"K5bf7kulHFISMqJ0uzBo12v9sb+fVeAsXH3FwOc0YH5B04kphgJsGWZe+PKiVZXXMzf3LJzQeGUHRg6AGsPIAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T03:27:21.979734Z"},"content_sha256":"ef7b296b77f6ea0986cd462842522a45673631362b916b1566845ab7aa53c755","schema_version":"1.0","event_id":"sha256:ef7b296b77f6ea0986cd462842522a45673631362b916b1566845ab7aa53c755"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:6M6GD7VD5G3YLJT6MSXERFVDKI","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"When Diffusion Breaks Constraints: Sequential Autoregressive Generation with RL and MCTS","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Diffusion models fail to satisfy strict geometric constraints in planning tasks because continuous density matching cannot target low-dimensional feasible regions, while reformulating generation as sequential discrete choices with RL and M-","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CV","authors_text":"Boye Niu, David Hsu, Harold Soh, Wee Sun Lee, Zirui Zhao","submitted_at":"2025-12-01T03:38:44Z","abstract_excerpt":"Data-driven generative models excel in language and vision, but diffusion models often fail in constrained planning and design tasks, exhibiting severe constraint violations in engineering inverse design, molecular generation, multi-robot planning, and floorplan/scene synthesis even with projection or guidance. Such tasks combine hard-to-specify semantic goals with strict geometric or physical constraints (e.g., non-overlap, connectivity), yielding feasible solutions that lie on low-dimensional, small, and sometimes disconnected regions of the output space. This paper studies the failure mode "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Overall, the empirical, theoretical, and prior-work evidence points to a structural limitation of continuous density matching on this class of constrained-generation problems, and suggests sequential constraint-aware generation as a promising alternative.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the failure modes observed in tangram generation from language and the simplified rectangle composition task are representative of the broader class of constrained planning and design tasks mentioned, including engineering inverse design, molecular generation, and multi-robot planning.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Diffusion models exhibit a structural limitation when generating samples on low-dimensional feasible regions for constrained tasks, and sequential autoregressive generation using RL and MCTS improves constraint satisfaction.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Diffusion models fail to satisfy strict geometric constraints in planning tasks because continuous density matching cannot target low-dimensional feasible regions, while reformulating generation as sequential discrete choices with RL and M-","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"b3b29e718046a1983174946088e1d9057edb19008d869141c6dd08b2f9b49366"},"source":{"id":"2512.01242","kind":"arxiv","version":3},"verdict":{"id":"6c9f7214-1b21-4cf1-81ba-593cef0e1aa7","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-17T03:40:48.393355Z","strongest_claim":"Overall, the empirical, theoretical, and prior-work evidence points to a structural limitation of continuous density matching on this class of constrained-generation problems, and suggests sequential constraint-aware generation as a promising alternative.","one_line_summary":"Diffusion models exhibit a structural limitation when generating samples on low-dimensional feasible regions for constrained tasks, and sequential autoregressive generation using RL and MCTS improves constraint satisfaction.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the failure modes observed in tangram generation from language and the simplified rectangle composition task are representative of the broader class of constrained planning and design tasks mentioned, including engineering inverse design, molecular generation, and multi-robot planning.","pith_extraction_headline":"Diffusion models fail to satisfy strict geometric constraints in planning tasks because continuous density matching cannot target low-dimensional feasible regions, while reformulating generation as sequential discrete choices with RL and M-"},"references":{"count":58,"sample":[{"doi":"","year":2010,"title":"Shape related constraints aware gen- eration of mechanical designs through deep convolutional gan.arXiv preprint arXiv:2010.11833, 2020","work_id":"2fafaa54-0b78-4ec3-aa4d-2575677737cb","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"Lexical entrainment without conceptual pacts? revisiting the matching task.Journal of Memory and Language, 114: 104129, 2020","work_id":"392fe622-e7b5-4cc0-82dc-aeab8f221b5f","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":1987,"title":"Recognition-by-components: A theory of human image understanding.Psychological Review, 94(2): 115–147, 1987","work_id":"fc69b3b0-dacb-4735-a93b-124bfbedc049","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Ab- stract concepts: External influences, internal constraints, and methodological issues.Psychological Research, 86(8): 2370–2388, 2022","work_id":"9976609d-2e1c-4467-9964-bf0ffb84e204","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Interac- tion promotes the adaptation of referential conventions to the communicative context.Cognitive science, 43(8):e12780,","work_id":"c041ef07-1e82-4bf1-95ca-3b5e56f18b5d","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":58,"snapshot_sha256":"540f4e59f61027313a1aa728a719fbdff90b45082086ee1221eda33d18776856","internal_anchors":2},"formal_canon":{"evidence_count":2,"snapshot_sha256":"c439a428b5e269b5818863db566120c0ed1a61632e3c058fc52cf0c98e48639e"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"6c9f7214-1b21-4cf1-81ba-593cef0e1aa7"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T03:09:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QzL8eQftJV7ORyMKCKzuHUt+J6bGxJ+2nSnTI2DYe+QP3kg7aR//ciFJIzUbmkCBh6No+vp3m5Ou6sVu//D1CQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T03:27:21.980823Z"},"content_sha256":"a09406708ba035c690465bb96040aaed6695f0b834485c7a5966bb53ce84b3de","schema_version":"1.0","event_id":"sha256:a09406708ba035c690465bb96040aaed6695f0b834485c7a5966bb53ce84b3de"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6M6GD7VD5G3YLJT6MSXERFVDKI/bundle.json","state_url":"https://pith.science/pith/6M6GD7VD5G3YLJT6MSXERFVDKI/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6M6GD7VD5G3YLJT6MSXERFVDKI/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":"2026-05-28T03:27:21Z","links":{"resolver":"https://pith.science/pith/6M6GD7VD5G3YLJT6MSXERFVDKI","bundle":"https://pith.science/pith/6M6GD7VD5G3YLJT6MSXERFVDKI/bundle.json","state":"https://pith.science/pith/6M6GD7VD5G3YLJT6MSXERFVDKI/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6M6GD7VD5G3YLJT6MSXERFVDKI/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:6M6GD7VD5G3YLJT6MSXERFVDKI","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":"eb931f34e2bbcfa6c060e198ba433d82a3a4f392d989f11a1a608b3d34f17d16","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-12-01T03:38:44Z","title_canon_sha256":"1ceea254e081438fe6352f9c9baff7939cc5b472ceb099fd6d34850737767a8c"},"schema_version":"1.0","source":{"id":"2512.01242","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2512.01242","created_at":"2026-05-18T03:09:32Z"},{"alias_kind":"arxiv_version","alias_value":"2512.01242v3","created_at":"2026-05-18T03:09:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.01242","created_at":"2026-05-18T03:09:32Z"},{"alias_kind":"pith_short_12","alias_value":"6M6GD7VD5G3Y","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"6M6GD7VD5G3YLJT6","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"6M6GD7VD","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:a09406708ba035c690465bb96040aaed6695f0b834485c7a5966bb53ce84b3de","target":"graph","created_at":"2026-05-18T03:09:32Z","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":"Overall, the empirical, theoretical, and prior-work evidence points to a structural limitation of continuous density matching on this class of constrained-generation problems, and suggests sequential constraint-aware generation as a promising alternative."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the failure modes observed in tangram generation from language and the simplified rectangle composition task are representative of the broader class of constrained planning and design tasks mentioned, including engineering inverse design, molecular generation, and multi-robot planning."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Diffusion models exhibit a structural limitation when generating samples on low-dimensional feasible regions for constrained tasks, and sequential autoregressive generation using RL and MCTS improves constraint satisfaction."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Diffusion models fail to satisfy strict geometric constraints in planning tasks because continuous density matching cannot target low-dimensional feasible regions, while reformulating generation as sequential discrete choices with RL and M-"}],"snapshot_sha256":"b3b29e718046a1983174946088e1d9057edb19008d869141c6dd08b2f9b49366"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"c439a428b5e269b5818863db566120c0ed1a61632e3c058fc52cf0c98e48639e"},"paper":{"abstract_excerpt":"Data-driven generative models excel in language and vision, but diffusion models often fail in constrained planning and design tasks, exhibiting severe constraint violations in engineering inverse design, molecular generation, multi-robot planning, and floorplan/scene synthesis even with projection or guidance. Such tasks combine hard-to-specify semantic goals with strict geometric or physical constraints (e.g., non-overlap, connectivity), yielding feasible solutions that lie on low-dimensional, small, and sometimes disconnected regions of the output space. This paper studies the failure mode ","authors_text":"Boye Niu, David Hsu, Harold Soh, Wee Sun Lee, Zirui Zhao","cross_cats":["cs.AI","cs.CL"],"headline":"Diffusion models fail to satisfy strict geometric constraints in planning tasks because continuous density matching cannot target low-dimensional feasible regions, while reformulating generation as sequential discrete choices with RL and M-","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-12-01T03:38:44Z","title":"When Diffusion Breaks Constraints: Sequential Autoregressive Generation with RL and MCTS"},"references":{"count":58,"internal_anchors":2,"resolved_work":58,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Shape related constraints aware gen- eration of mechanical designs through deep convolutional gan.arXiv preprint arXiv:2010.11833, 2020","work_id":"2fafaa54-0b78-4ec3-aa4d-2575677737cb","year":2010},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Lexical entrainment without conceptual pacts? revisiting the matching task.Journal of Memory and Language, 114: 104129, 2020","work_id":"392fe622-e7b5-4cc0-82dc-aeab8f221b5f","year":2020},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Recognition-by-components: A theory of human image understanding.Psychological Review, 94(2): 115–147, 1987","work_id":"fc69b3b0-dacb-4735-a93b-124bfbedc049","year":1987},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Ab- stract concepts: External influences, internal constraints, and methodological issues.Psychological Research, 86(8): 2370–2388, 2022","work_id":"9976609d-2e1c-4467-9964-bf0ffb84e204","year":2022},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Interac- tion promotes the adaptation of referential conventions to the communicative context.Cognitive science, 43(8):e12780,","work_id":"c041ef07-1e82-4bf1-95ca-3b5e56f18b5d","year":null}],"snapshot_sha256":"540f4e59f61027313a1aa728a719fbdff90b45082086ee1221eda33d18776856"},"source":{"id":"2512.01242","kind":"arxiv","version":3},"verdict":{"created_at":"2026-05-17T03:40:48.393355Z","id":"6c9f7214-1b21-4cf1-81ba-593cef0e1aa7","model_set":{"reader":"grok-4.3"},"one_line_summary":"Diffusion models exhibit a structural limitation when generating samples on low-dimensional feasible regions for constrained tasks, and sequential autoregressive generation using RL and MCTS improves constraint satisfaction.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Diffusion models fail to satisfy strict geometric constraints in planning tasks because continuous density matching cannot target low-dimensional feasible regions, while reformulating generation as sequential discrete choices with RL and M-","strongest_claim":"Overall, the empirical, theoretical, and prior-work evidence points to a structural limitation of continuous density matching on this class of constrained-generation problems, and suggests sequential constraint-aware generation as a promising alternative.","weakest_assumption":"That the failure modes observed in tangram generation from language and the simplified rectangle composition task are representative of the broader class of constrained planning and design tasks mentioned, including engineering inverse design, molecular generation, and multi-robot planning."}},"verdict_id":"6c9f7214-1b21-4cf1-81ba-593cef0e1aa7"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:ef7b296b77f6ea0986cd462842522a45673631362b916b1566845ab7aa53c755","target":"record","created_at":"2026-05-18T03:09:32Z","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":"eb931f34e2bbcfa6c060e198ba433d82a3a4f392d989f11a1a608b3d34f17d16","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-12-01T03:38:44Z","title_canon_sha256":"1ceea254e081438fe6352f9c9baff7939cc5b472ceb099fd6d34850737767a8c"},"schema_version":"1.0","source":{"id":"2512.01242","kind":"arxiv","version":3}},"canonical_sha256":"f33c61fea3e9b785a67e64ae4896a3521a79c5e4b1e62f431599d2212833c4b9","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f33c61fea3e9b785a67e64ae4896a3521a79c5e4b1e62f431599d2212833c4b9","first_computed_at":"2026-05-18T03:09:32.924055Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T03:09:32.924055Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"aMg31aBYCXfrK+uNGnnQXuL4IWdTh7tIPPAECnLpnmez2MKLF/N8/EzuzcGjy4+oPwIQJVgKyU2UkZNwnptFCQ==","signature_status":"signed_v1","signed_at":"2026-05-18T03:09:32.924798Z","signed_message":"canonical_sha256_bytes"},"source_id":"2512.01242","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ef7b296b77f6ea0986cd462842522a45673631362b916b1566845ab7aa53c755","sha256:a09406708ba035c690465bb96040aaed6695f0b834485c7a5966bb53ce84b3de"],"state_sha256":"03cbc4710c54eb10dbf4712971c2ab53538c15a3084baab66c99ea483b54faf6"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oYV1U7z9BXrYY7/M4PXFTP/27JNpgnb81kjxXqUHL5EgbRtOYpiM0Ggyl0r15kg7TtAadWKYHkUvoZzza7PGCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-28T03:27:21.984868Z","bundle_sha256":"05bf4d5d4a6d057e95ce806fe40be4c07e14a6e68fac81e88ab959dae9357f80"}}