{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:6RE6OQXOC7VBYN5JTT2AZ2BXUR","short_pith_number":"pith:6RE6OQXO","canonical_record":{"source":{"id":"2603.20334","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-03-20T08:16:15Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c18d5ca1e01f3a5b5ed7f345e467d8806a09cef426817b50fc3c173fdfb161e2","abstract_canon_sha256":"517fc8e3d3e2fd84cc027c1a5e50bd14da025336d1bfdd24e030b9a53522fe61"},"schema_version":"1.0"},"canonical_sha256":"f449e742ee17ea1c37a99cf40ce837a455b7801197d8c45e4c50cbe6e7d1d7c4","source":{"kind":"arxiv","id":"2603.20334","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.20334","created_at":"2026-05-18T02:45:04Z"},{"alias_kind":"arxiv_version","alias_value":"2603.20334v3","created_at":"2026-05-18T02:45:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.20334","created_at":"2026-05-18T02:45:04Z"},{"alias_kind":"pith_short_12","alias_value":"6RE6OQXOC7VB","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"6RE6OQXOC7VBYN5J","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"6RE6OQXO","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:6RE6OQXOC7VBYN5JTT2AZ2BXUR","target":"record","payload":{"canonical_record":{"source":{"id":"2603.20334","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-03-20T08:16:15Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c18d5ca1e01f3a5b5ed7f345e467d8806a09cef426817b50fc3c173fdfb161e2","abstract_canon_sha256":"517fc8e3d3e2fd84cc027c1a5e50bd14da025336d1bfdd24e030b9a53522fe61"},"schema_version":"1.0"},"canonical_sha256":"f449e742ee17ea1c37a99cf40ce837a455b7801197d8c45e4c50cbe6e7d1d7c4","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:45:04.711908Z","signature_b64":"cLKfs8lTJxw/tNfL392vxAcxXYlUFhPaSt9K+0ExHxiXsvdYbaJDOsrxWrQue6TJz9gVd921xJeX7/PmMFABCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f449e742ee17ea1c37a99cf40ce837a455b7801197d8c45e4c50cbe6e7d1d7c4","last_reissued_at":"2026-05-18T02:45:04.711443Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:45:04.711443Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2603.20334","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":"2026-05-18T02:45:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Sq68uvlLz2jGVtass71oUth3esv101ulya9GvFsWseHc6qe0854a8KvCNhkFy9MDItqP640++NztrvN7+m9/Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-02T07:06:09.338089Z"},"content_sha256":"ca4d477cc3042f881de1c352922cfed31175fecdad6a00a5b88d8e7451f302e2","schema_version":"1.0","event_id":"sha256:ca4d477cc3042f881de1c352922cfed31175fecdad6a00a5b88d8e7451f302e2"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:6RE6OQXOC7VBYN5JTT2AZ2BXUR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Procedural Refinement by LLM-driven Algorithmic Debugging for ARC-AGI-2","license":"http://creativecommons.org/licenses/by/4.0/","headline":"ABPR couples LLMs with a Prolog meta-interpreter to debug hypothesized transformation rules through SLD proof trees on ARC-AGI-2 tasks.","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Jiong-Da Wang, Lin-Feng Zou, Wang-Zhou Dai, Xue-Rong Yuan, Yu-Ning Qiu","submitted_at":"2026-03-20T08:16:15Z","abstract_excerpt":"In high-complexity abstract reasoning, a system must infer a latent rule from a few examples or structured observations and apply it to unseen instances. LLMs can express such rules as programs, but ordinary conversation-based refinement is largely outcome-level: it observes that an answer or output is wrong without formally re-checking which abstraction, relation, or transformation justified that outcome. We propose \\emph{Abduction-Based Procedural Refinement} (ABPR), a neuro-symbolic refinement approach that couples an LLM with a Prolog meta-interpreter. ABPR treats each candidate program as"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"ABPR with Gemini-3-Flash achieves 56.67% Pass@2, while GPT-5.5 xHigh with ABPR reaches 98.33% Pass@2 on the public evaluation set of ARC-AGI-2.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the Prolog meta-interpreter can faithfully reify the LLM's hypothesized transformation rules into SLD proof trees whose subgoals correspond to the actual latent abstractions needed for grid and relational tasks.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"ABPR uses LLM-generated programs debugged through Prolog SLD proof traces to reach 56.67% Pass@2 with Gemini-3-Flash and 98.33% with GPT-5.5 xHigh on ARC-AGI-2.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"ABPR couples LLMs with a Prolog meta-interpreter to debug hypothesized transformation rules through SLD proof trees on ARC-AGI-2 tasks.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"defef3b17a116b4833511a54627c8d4738ebb9ee641d4c7bad6238fc59f59448"},"source":{"id":"2603.20334","kind":"arxiv","version":3},"verdict":{"id":"326d053c-9871-445a-b205-1a419ef04b16","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T09:02:57.510013Z","strongest_claim":"ABPR with Gemini-3-Flash achieves 56.67% Pass@2, while GPT-5.5 xHigh with ABPR reaches 98.33% Pass@2 on the public evaluation set of ARC-AGI-2.","one_line_summary":"ABPR uses LLM-generated programs debugged through Prolog SLD proof traces to reach 56.67% Pass@2 with Gemini-3-Flash and 98.33% with GPT-5.5 xHigh on ARC-AGI-2.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the Prolog meta-interpreter can faithfully reify the LLM's hypothesized transformation rules into SLD proof trees whose subgoals correspond to the actual latent abstractions needed for grid and relational tasks.","pith_extraction_headline":"ABPR couples LLMs with a Prolog meta-interpreter to debug hypothesized transformation rules through SLD proof trees on ARC-AGI-2 tasks."},"references":{"count":25,"sample":[{"doi":"","year":2022,"title":"The Curious Case of Neural Text Degeneration","work_id":"1ef2ec4a-db7b-42b2-9416-0f1bb628c3c8","ref_index":1,"cited_arxiv_id":"1904.09751","is_internal_anchor":true},{"doi":"10.48550/arxiv","year":2025,"title":"URLhttps://doi.org/10.48550/arXiv","work_id":"5c2060c6-427c-4321-be22-49ccae439d80","ref_index":2,"cited_arxiv_id":"2203.14987","is_internal_anchor":true},{"doi":"","year":null,"title":"**Training Examples ** - Input-output pairs to help you understand the transformation pattern","work_id":"4e0faea4-2e33-444f-bc12-c6de9577181e","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Your REAL GOAL is to produce a ‘solve/2‘ predicate that correctly transforms the Challenge input(s)","work_id":"219c00f6-c6ac-4fe5-9aad-c3d415bdc82f","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Note their dimensions, color palettes, and any prominent visual features (shapes, symmetries, patterns)","work_id":"0f4399eb-1dcb-41a0-bd9c-7a7bb213aa9e","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":25,"snapshot_sha256":"9c2a275a8766829f8ffcd21139ed817adff9b93d19fb92ba1ab1d33c31fc02f4","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":"326d053c-9871-445a-b205-1a419ef04b16"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:45:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5BOz9pHB3Y14DKTBb0hrAIu6G33i8jRHFDWhF5LbNHNim6S7gWanjhMux/uLjA6IYKqHKELxxLG2Ur489cEHCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-02T07:06:09.338737Z"},"content_sha256":"dc73dadd7684602ed78551191ea14a1396d22e8732d52247034c0b83a82a5d46","schema_version":"1.0","event_id":"sha256:dc73dadd7684602ed78551191ea14a1396d22e8732d52247034c0b83a82a5d46"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6RE6OQXOC7VBYN5JTT2AZ2BXUR/bundle.json","state_url":"https://pith.science/pith/6RE6OQXOC7VBYN5JTT2AZ2BXUR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6RE6OQXOC7VBYN5JTT2AZ2BXUR/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":"2026-06-02T07:06:09Z","links":{"resolver":"https://pith.science/pith/6RE6OQXOC7VBYN5JTT2AZ2BXUR","bundle":"https://pith.science/pith/6RE6OQXOC7VBYN5JTT2AZ2BXUR/bundle.json","state":"https://pith.science/pith/6RE6OQXOC7VBYN5JTT2AZ2BXUR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6RE6OQXOC7VBYN5JTT2AZ2BXUR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:6RE6OQXOC7VBYN5JTT2AZ2BXUR","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":"517fc8e3d3e2fd84cc027c1a5e50bd14da025336d1bfdd24e030b9a53522fe61","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-03-20T08:16:15Z","title_canon_sha256":"c18d5ca1e01f3a5b5ed7f345e467d8806a09cef426817b50fc3c173fdfb161e2"},"schema_version":"1.0","source":{"id":"2603.20334","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.20334","created_at":"2026-05-18T02:45:04Z"},{"alias_kind":"arxiv_version","alias_value":"2603.20334v3","created_at":"2026-05-18T02:45:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.20334","created_at":"2026-05-18T02:45:04Z"},{"alias_kind":"pith_short_12","alias_value":"6RE6OQXOC7VB","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"6RE6OQXOC7VBYN5J","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"6RE6OQXO","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:dc73dadd7684602ed78551191ea14a1396d22e8732d52247034c0b83a82a5d46","target":"graph","created_at":"2026-05-18T02:45:04Z","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":"ABPR with Gemini-3-Flash achieves 56.67% Pass@2, while GPT-5.5 xHigh with ABPR reaches 98.33% Pass@2 on the public evaluation set of ARC-AGI-2."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the Prolog meta-interpreter can faithfully reify the LLM's hypothesized transformation rules into SLD proof trees whose subgoals correspond to the actual latent abstractions needed for grid and relational tasks."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"ABPR uses LLM-generated programs debugged through Prolog SLD proof traces to reach 56.67% Pass@2 with Gemini-3-Flash and 98.33% with GPT-5.5 xHigh on ARC-AGI-2."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"ABPR couples LLMs with a Prolog meta-interpreter to debug hypothesized transformation rules through SLD proof trees on ARC-AGI-2 tasks."}],"snapshot_sha256":"defef3b17a116b4833511a54627c8d4738ebb9ee641d4c7bad6238fc59f59448"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"In high-complexity abstract reasoning, a system must infer a latent rule from a few examples or structured observations and apply it to unseen instances. LLMs can express such rules as programs, but ordinary conversation-based refinement is largely outcome-level: it observes that an answer or output is wrong without formally re-checking which abstraction, relation, or transformation justified that outcome. We propose \\emph{Abduction-Based Procedural Refinement} (ABPR), a neuro-symbolic refinement approach that couples an LLM with a Prolog meta-interpreter. ABPR treats each candidate program as","authors_text":"Jiong-Da Wang, Lin-Feng Zou, Wang-Zhou Dai, Xue-Rong Yuan, Yu-Ning Qiu","cross_cats":["cs.AI"],"headline":"ABPR couples LLMs with a Prolog meta-interpreter to debug hypothesized transformation rules through SLD proof trees on ARC-AGI-2 tasks.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-03-20T08:16:15Z","title":"Procedural Refinement by LLM-driven Algorithmic Debugging for ARC-AGI-2"},"references":{"count":25,"internal_anchors":2,"resolved_work":25,"sample":[{"cited_arxiv_id":"1904.09751","doi":"","is_internal_anchor":true,"ref_index":1,"title":"The Curious Case of Neural Text Degeneration","work_id":"1ef2ec4a-db7b-42b2-9416-0f1bb628c3c8","year":2022},{"cited_arxiv_id":"2203.14987","doi":"10.48550/arxiv","is_internal_anchor":true,"ref_index":2,"title":"URLhttps://doi.org/10.48550/arXiv","work_id":"5c2060c6-427c-4321-be22-49ccae439d80","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"**Training Examples ** - Input-output pairs to help you understand the transformation pattern","work_id":"4e0faea4-2e33-444f-bc12-c6de9577181e","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Your REAL GOAL is to produce a ‘solve/2‘ predicate that correctly transforms the Challenge input(s)","work_id":"219c00f6-c6ac-4fe5-9aad-c3d415bdc82f","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Note their dimensions, color palettes, and any prominent visual features (shapes, symmetries, patterns)","work_id":"0f4399eb-1dcb-41a0-bd9c-7a7bb213aa9e","year":null}],"snapshot_sha256":"9c2a275a8766829f8ffcd21139ed817adff9b93d19fb92ba1ab1d33c31fc02f4"},"source":{"id":"2603.20334","kind":"arxiv","version":3},"verdict":{"created_at":"2026-05-15T09:02:57.510013Z","id":"326d053c-9871-445a-b205-1a419ef04b16","model_set":{"reader":"grok-4.3"},"one_line_summary":"ABPR uses LLM-generated programs debugged through Prolog SLD proof traces to reach 56.67% Pass@2 with Gemini-3-Flash and 98.33% with GPT-5.5 xHigh on ARC-AGI-2.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"ABPR couples LLMs with a Prolog meta-interpreter to debug hypothesized transformation rules through SLD proof trees on ARC-AGI-2 tasks.","strongest_claim":"ABPR with Gemini-3-Flash achieves 56.67% Pass@2, while GPT-5.5 xHigh with ABPR reaches 98.33% Pass@2 on the public evaluation set of ARC-AGI-2.","weakest_assumption":"That the Prolog meta-interpreter can faithfully reify the LLM's hypothesized transformation rules into SLD proof trees whose subgoals correspond to the actual latent abstractions needed for grid and relational tasks."}},"verdict_id":"326d053c-9871-445a-b205-1a419ef04b16"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:ca4d477cc3042f881de1c352922cfed31175fecdad6a00a5b88d8e7451f302e2","target":"record","created_at":"2026-05-18T02:45:04Z","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":"517fc8e3d3e2fd84cc027c1a5e50bd14da025336d1bfdd24e030b9a53522fe61","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-03-20T08:16:15Z","title_canon_sha256":"c18d5ca1e01f3a5b5ed7f345e467d8806a09cef426817b50fc3c173fdfb161e2"},"schema_version":"1.0","source":{"id":"2603.20334","kind":"arxiv","version":3}},"canonical_sha256":"f449e742ee17ea1c37a99cf40ce837a455b7801197d8c45e4c50cbe6e7d1d7c4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f449e742ee17ea1c37a99cf40ce837a455b7801197d8c45e4c50cbe6e7d1d7c4","first_computed_at":"2026-05-18T02:45:04.711443Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T02:45:04.711443Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"cLKfs8lTJxw/tNfL392vxAcxXYlUFhPaSt9K+0ExHxiXsvdYbaJDOsrxWrQue6TJz9gVd921xJeX7/PmMFABCA==","signature_status":"signed_v1","signed_at":"2026-05-18T02:45:04.711908Z","signed_message":"canonical_sha256_bytes"},"source_id":"2603.20334","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ca4d477cc3042f881de1c352922cfed31175fecdad6a00a5b88d8e7451f302e2","sha256:dc73dadd7684602ed78551191ea14a1396d22e8732d52247034c0b83a82a5d46"],"state_sha256":"8a185ff9a1b9f0ca9ce91388dea27da525995b1da60d4f2e7ef877f35d8be315"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OvXXPZG1ekrfGbvQvSE1vkZWMCcKCLzHNVxqOyBDkurJdvpiwza26f102xCIGyMWiBBZYPomtMHYU06e5BEKCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-02T07:06:09.341747Z","bundle_sha256":"0893b204609ecbd9fa7b6aff538814640061d7f57f01f6afdfc69c0471b77cb4"}}