{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:BYFFEMDJN5AAXFGZ5K2OHHRJ44","short_pith_number":"pith:BYFFEMDJ","canonical_record":{"source":{"id":"2605.16842","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-16T06:59:54Z","cross_cats_sorted":[],"title_canon_sha256":"58a4888ac1e2fff49fc3abc3084a79e22f0b0c689af7db46ae3d44292e3f65a7","abstract_canon_sha256":"5b1c3ab3f01a8eda0f7c556aeac57b2b5fe7414e73ae723e5628244892489f00"},"schema_version":"1.0"},"canonical_sha256":"0e0a5230696f400b94d9eab4e39e29e71158dca69d8cdf346aab34f3a1460273","source":{"kind":"arxiv","id":"2605.16842","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.16842","created_at":"2026-05-20T00:03:25Z"},{"alias_kind":"arxiv_version","alias_value":"2605.16842v1","created_at":"2026-05-20T00:03:25Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.16842","created_at":"2026-05-20T00:03:25Z"},{"alias_kind":"pith_short_12","alias_value":"BYFFEMDJN5AA","created_at":"2026-05-20T00:03:25Z"},{"alias_kind":"pith_short_16","alias_value":"BYFFEMDJN5AAXFGZ","created_at":"2026-05-20T00:03:25Z"},{"alias_kind":"pith_short_8","alias_value":"BYFFEMDJ","created_at":"2026-05-20T00:03:25Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:BYFFEMDJN5AAXFGZ5K2OHHRJ44","target":"record","payload":{"canonical_record":{"source":{"id":"2605.16842","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-16T06:59:54Z","cross_cats_sorted":[],"title_canon_sha256":"58a4888ac1e2fff49fc3abc3084a79e22f0b0c689af7db46ae3d44292e3f65a7","abstract_canon_sha256":"5b1c3ab3f01a8eda0f7c556aeac57b2b5fe7414e73ae723e5628244892489f00"},"schema_version":"1.0"},"canonical_sha256":"0e0a5230696f400b94d9eab4e39e29e71158dca69d8cdf346aab34f3a1460273","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:03:25.756105Z","signature_b64":"vnDyccNZCgVmITNLTO6/KJm0btMPS0tjz2oYBJtAJRvalhAyjczHEcSqVPlexcrZzoFX8eZ6aDgwMfynmtVFAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0e0a5230696f400b94d9eab4e39e29e71158dca69d8cdf346aab34f3a1460273","last_reissued_at":"2026-05-20T00:03:25.755287Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:03:25.755287Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.16842","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":"2026-05-20T00:03:25Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"+bVhvN7g2sI1hjwNZA6ZFaNuUT18M8++YmjtoVcv2YkYTzDvrDg/ersZ2a/3ZqlCm8Jf5szpI2J/GgANC75kCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T03:04:43.331490Z"},"content_sha256":"700ea2569cded2c1714e2b9ccdf2950772ed1c29b9561d6bf1af1a25b857fb72","schema_version":"1.0","event_id":"sha256:700ea2569cded2c1714e2b9ccdf2950772ed1c29b9561d6bf1af1a25b857fb72"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:BYFFEMDJN5AAXFGZ5K2OHHRJ44","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Sketch Then Paint: Hierarchical Reinforcement Learning for Diffusion Multi-Modal Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Hierarchical reinforcement learning with staged sketch-then-paint updates improves how diffusion multimodal models assign credit during image generation.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Guangtao Zhai, Haoxing Chen, Huayu Zheng, Jianghan Shen, Junjun He, Siqi Luo, Xiaohong Liu, Yan Tai, Yihao Liu, Yi Xin, Yue Li, Yuewen Cao","submitted_at":"2026-05-16T06:59:54Z","abstract_excerpt":"Diffusion Multi-Modal Large Language Models (dMLLMs) are powerful for image generation, but optimizing them through reinforcement learning (RL) remains a major challenge. One primary difficulty is that a single image can be generated through many different unmasking sequences, which makes calculating importance ratios often intractable. Additionally, existing methods tend to ignore the hierarchical generation process of dMLLMs, where early tokens define the global layout and later tokens focus on local details. By assigning uniform rewards to all tokens, these current methods fail to reflect t"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"HT-GRPO achieves substantial gains on the GenEval and DPG benchmarks. Evaluations across six additional metrics confirm significant improvements in image quality, aesthetics, and human preference.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The prompt-conditioned estimator correctly computes importance ratios from a fully masked state and the hierarchical credit assignment accurately prioritizes structural tokens without introducing new biases or instabilities in the policy optimization.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Proposes HT-GRPO with sketch-then-paint staged updates, prompt-conditioned importance ratios, and hierarchical credit assignment for dMLLMs, reporting gains on GenEval and DPG plus quality metrics.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Hierarchical reinforcement learning with staged sketch-then-paint updates improves how diffusion multimodal models assign credit during image generation.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"b5e20f0492484be15fee3f59dcbe91bc698f0f0a95adeae75087099c6a932673"},"source":{"id":"2605.16842","kind":"arxiv","version":1},"verdict":{"id":"f0eaa786-d9a7-4adb-bf21-e585e474ac81","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T21:14:19.995559Z","strongest_claim":"HT-GRPO achieves substantial gains on the GenEval and DPG benchmarks. Evaluations across six additional metrics confirm significant improvements in image quality, aesthetics, and human preference.","one_line_summary":"Proposes HT-GRPO with sketch-then-paint staged updates, prompt-conditioned importance ratios, and hierarchical credit assignment for dMLLMs, reporting gains on GenEval and DPG plus quality metrics.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The prompt-conditioned estimator correctly computes importance ratios from a fully masked state and the hierarchical credit assignment accurately prioritizes structural tokens without introducing new biases or instabilities in the policy optimization.","pith_extraction_headline":"Hierarchical reinforcement learning with staged sketch-then-paint updates improves how diffusion multimodal models assign credit during image generation."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.16842/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T21:31:19.231483Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T21:20:56.717213Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T18:41:56.320997Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T18:33:26.393755Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"12ea26043c921702b4a2986e5bfb78fbcaca9906ff4fb815f56ec9eefdbb51c1"},"references":{"count":49,"sample":[{"doi":"","year":2025,"title":"Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer","work_id":"f1080a62-48e1-4255-b023-7556be57370d","ref_index":1,"cited_arxiv_id":"2511.22699","is_internal_anchor":true},{"doi":"","year":2025,"title":"Lumina-image 2.0: A unified and efficient image generative framework","work_id":"e8f43e3f-d074-4791-b07c-3ac843c39e30","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Qwen-Image Technical Report","work_id":"d06d7ecc-7579-4f89-a60b-4278a0f3c562","ref_index":3,"cited_arxiv_id":"2508.02324","is_internal_anchor":true},{"doi":"","year":2025,"title":"LongCat-Image Technical Report","work_id":"648d9514-ca31-4ba0-983a-1aba1057a97e","ref_index":4,"cited_arxiv_id":"2512.07584","is_internal_anchor":true},{"doi":"","year":2024,"title":"Emu3: Next-Token Prediction is All You Need","work_id":"720d288e-fac0-464c-9929-19efd9a52afc","ref_index":5,"cited_arxiv_id":"2409.18869","is_internal_anchor":true}],"resolved_work":49,"snapshot_sha256":"12bd7db934788fcb64531d3b2933c72e6e2db323d74a96e0c07cc93374924385","internal_anchors":18},"formal_canon":{"evidence_count":2,"snapshot_sha256":"97a852457b151b2f34ef814c1ebccc8cd8495ede5be7e585fc0e8189e7aa0b9a"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"f0eaa786-d9a7-4adb-bf21-e585e474ac81"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T00:03:25Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0dd6gHb4IO+QHc1qvYajImyu8ViKfSBQHQHs3ppwWLUg9rhEai1I3mdVY4XahU6NnV9x2s3OF8e6PvKL2QMgBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T03:04:43.332681Z"},"content_sha256":"990ddc6b62cc4f106bc7dbe6728352c3e01f8defb31516e78a21e760e9366790","schema_version":"1.0","event_id":"sha256:990ddc6b62cc4f106bc7dbe6728352c3e01f8defb31516e78a21e760e9366790"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/BYFFEMDJN5AAXFGZ5K2OHHRJ44/bundle.json","state_url":"https://pith.science/pith/BYFFEMDJN5AAXFGZ5K2OHHRJ44/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/BYFFEMDJN5AAXFGZ5K2OHHRJ44/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":"2026-05-25T03:04:43Z","links":{"resolver":"https://pith.science/pith/BYFFEMDJN5AAXFGZ5K2OHHRJ44","bundle":"https://pith.science/pith/BYFFEMDJN5AAXFGZ5K2OHHRJ44/bundle.json","state":"https://pith.science/pith/BYFFEMDJN5AAXFGZ5K2OHHRJ44/state.json","well_known_bundle":"https://pith.science/.well-known/pith/BYFFEMDJN5AAXFGZ5K2OHHRJ44/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:BYFFEMDJN5AAXFGZ5K2OHHRJ44","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":"5b1c3ab3f01a8eda0f7c556aeac57b2b5fe7414e73ae723e5628244892489f00","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-16T06:59:54Z","title_canon_sha256":"58a4888ac1e2fff49fc3abc3084a79e22f0b0c689af7db46ae3d44292e3f65a7"},"schema_version":"1.0","source":{"id":"2605.16842","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.16842","created_at":"2026-05-20T00:03:25Z"},{"alias_kind":"arxiv_version","alias_value":"2605.16842v1","created_at":"2026-05-20T00:03:25Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.16842","created_at":"2026-05-20T00:03:25Z"},{"alias_kind":"pith_short_12","alias_value":"BYFFEMDJN5AA","created_at":"2026-05-20T00:03:25Z"},{"alias_kind":"pith_short_16","alias_value":"BYFFEMDJN5AAXFGZ","created_at":"2026-05-20T00:03:25Z"},{"alias_kind":"pith_short_8","alias_value":"BYFFEMDJ","created_at":"2026-05-20T00:03:25Z"}],"graph_snapshots":[{"event_id":"sha256:990ddc6b62cc4f106bc7dbe6728352c3e01f8defb31516e78a21e760e9366790","target":"graph","created_at":"2026-05-20T00:03:25Z","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":"HT-GRPO achieves substantial gains on the GenEval and DPG benchmarks. Evaluations across six additional metrics confirm significant improvements in image quality, aesthetics, and human preference."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The prompt-conditioned estimator correctly computes importance ratios from a fully masked state and the hierarchical credit assignment accurately prioritizes structural tokens without introducing new biases or instabilities in the policy optimization."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Proposes HT-GRPO with sketch-then-paint staged updates, prompt-conditioned importance ratios, and hierarchical credit assignment for dMLLMs, reporting gains on GenEval and DPG plus quality metrics."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Hierarchical reinforcement learning with staged sketch-then-paint updates improves how diffusion multimodal models assign credit during image generation."}],"snapshot_sha256":"b5e20f0492484be15fee3f59dcbe91bc698f0f0a95adeae75087099c6a932673"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"97a852457b151b2f34ef814c1ebccc8cd8495ede5be7e585fc0e8189e7aa0b9a"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"doi_title_agreement","ran_at":"2026-05-19T21:31:19.231483Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T21:20:56.717213Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"claim_evidence","ran_at":"2026-05-19T18:41:56.320997Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-19T18:33:26.393755Z","status":"skipped","version":"1.0.0"}],"endpoint":"/pith/2605.16842/integrity.json","findings":[],"snapshot_sha256":"12ea26043c921702b4a2986e5bfb78fbcaca9906ff4fb815f56ec9eefdbb51c1","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Diffusion Multi-Modal Large Language Models (dMLLMs) are powerful for image generation, but optimizing them through reinforcement learning (RL) remains a major challenge. One primary difficulty is that a single image can be generated through many different unmasking sequences, which makes calculating importance ratios often intractable. Additionally, existing methods tend to ignore the hierarchical generation process of dMLLMs, where early tokens define the global layout and later tokens focus on local details. By assigning uniform rewards to all tokens, these current methods fail to reflect t","authors_text":"Guangtao Zhai, Haoxing Chen, Huayu Zheng, Jianghan Shen, Junjun He, Siqi Luo, Xiaohong Liu, Yan Tai, Yihao Liu, Yi Xin, Yue Li, Yuewen Cao","cross_cats":[],"headline":"Hierarchical reinforcement learning with staged sketch-then-paint updates improves how diffusion multimodal models assign credit during image generation.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-16T06:59:54Z","title":"Sketch Then Paint: Hierarchical Reinforcement Learning for Diffusion Multi-Modal Large Language Models"},"references":{"count":49,"internal_anchors":18,"resolved_work":49,"sample":[{"cited_arxiv_id":"2511.22699","doi":"","is_internal_anchor":true,"ref_index":1,"title":"Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer","work_id":"f1080a62-48e1-4255-b023-7556be57370d","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Lumina-image 2.0: A unified and efficient image generative framework","work_id":"e8f43e3f-d074-4791-b07c-3ac843c39e30","year":2025},{"cited_arxiv_id":"2508.02324","doi":"","is_internal_anchor":true,"ref_index":3,"title":"Qwen-Image Technical Report","work_id":"d06d7ecc-7579-4f89-a60b-4278a0f3c562","year":2025},{"cited_arxiv_id":"2512.07584","doi":"","is_internal_anchor":true,"ref_index":4,"title":"LongCat-Image Technical Report","work_id":"648d9514-ca31-4ba0-983a-1aba1057a97e","year":2025},{"cited_arxiv_id":"2409.18869","doi":"","is_internal_anchor":true,"ref_index":5,"title":"Emu3: Next-Token Prediction is All You Need","work_id":"720d288e-fac0-464c-9929-19efd9a52afc","year":2024}],"snapshot_sha256":"12bd7db934788fcb64531d3b2933c72e6e2db323d74a96e0c07cc93374924385"},"source":{"id":"2605.16842","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-19T21:14:19.995559Z","id":"f0eaa786-d9a7-4adb-bf21-e585e474ac81","model_set":{"reader":"grok-4.3"},"one_line_summary":"Proposes HT-GRPO with sketch-then-paint staged updates, prompt-conditioned importance ratios, and hierarchical credit assignment for dMLLMs, reporting gains on GenEval and DPG plus quality metrics.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Hierarchical reinforcement learning with staged sketch-then-paint updates improves how diffusion multimodal models assign credit during image generation.","strongest_claim":"HT-GRPO achieves substantial gains on the GenEval and DPG benchmarks. Evaluations across six additional metrics confirm significant improvements in image quality, aesthetics, and human preference.","weakest_assumption":"The prompt-conditioned estimator correctly computes importance ratios from a fully masked state and the hierarchical credit assignment accurately prioritizes structural tokens without introducing new biases or instabilities in the policy optimization."}},"verdict_id":"f0eaa786-d9a7-4adb-bf21-e585e474ac81"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:700ea2569cded2c1714e2b9ccdf2950772ed1c29b9561d6bf1af1a25b857fb72","target":"record","created_at":"2026-05-20T00:03:25Z","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":"5b1c3ab3f01a8eda0f7c556aeac57b2b5fe7414e73ae723e5628244892489f00","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-16T06:59:54Z","title_canon_sha256":"58a4888ac1e2fff49fc3abc3084a79e22f0b0c689af7db46ae3d44292e3f65a7"},"schema_version":"1.0","source":{"id":"2605.16842","kind":"arxiv","version":1}},"canonical_sha256":"0e0a5230696f400b94d9eab4e39e29e71158dca69d8cdf346aab34f3a1460273","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0e0a5230696f400b94d9eab4e39e29e71158dca69d8cdf346aab34f3a1460273","first_computed_at":"2026-05-20T00:03:25.755287Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:03:25.755287Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"vnDyccNZCgVmITNLTO6/KJm0btMPS0tjz2oYBJtAJRvalhAyjczHEcSqVPlexcrZzoFX8eZ6aDgwMfynmtVFAw==","signature_status":"signed_v1","signed_at":"2026-05-20T00:03:25.756105Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.16842","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:700ea2569cded2c1714e2b9ccdf2950772ed1c29b9561d6bf1af1a25b857fb72","sha256:990ddc6b62cc4f106bc7dbe6728352c3e01f8defb31516e78a21e760e9366790"],"state_sha256":"ec497fc6a7a6d363940bac3fd5fd459fb107a385b8783c993f4014d36064cf85"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KZIn2NWpLZisolfv/X9BhzFWpEn+gMNOVZMgZBQqyJ6jaKlpGuTnw3zNVZ7waE5O4k1U61/Zy5CHDnXCg6jCBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-25T03:04:43.337591Z","bundle_sha256":"435c8ab0853949ca8da6bcd783db9a818928ccbad46f9fe8fcf5383c123662a7"}}