{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:R73BLH45A2YJVPTU4VOOZ6R6VB","short_pith_number":"pith:R73BLH45","canonical_record":{"source":{"id":"2605.08949","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-09T13:42:08Z","cross_cats_sorted":[],"title_canon_sha256":"3f0a057ff859e968774069b12909d97b2786e1f57687776717abc66d0785e963","abstract_canon_sha256":"e41baa924b7c84a8c651311634366eb397c03503c36920ec0ac0e42a78e490f7"},"schema_version":"1.0"},"canonical_sha256":"8ff6159f9d06b09abe74e55cecfa3ea8500f9575464183abd463ef92835a343a","source":{"kind":"arxiv","id":"2605.08949","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.08949","created_at":"2026-05-20T00:00:41Z"},{"alias_kind":"arxiv_version","alias_value":"2605.08949v2","created_at":"2026-05-20T00:00:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.08949","created_at":"2026-05-20T00:00:41Z"},{"alias_kind":"pith_short_12","alias_value":"R73BLH45A2YJ","created_at":"2026-05-20T00:00:41Z"},{"alias_kind":"pith_short_16","alias_value":"R73BLH45A2YJVPTU","created_at":"2026-05-20T00:00:41Z"},{"alias_kind":"pith_short_8","alias_value":"R73BLH45","created_at":"2026-05-20T00:00:41Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:R73BLH45A2YJVPTU4VOOZ6R6VB","target":"record","payload":{"canonical_record":{"source":{"id":"2605.08949","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-09T13:42:08Z","cross_cats_sorted":[],"title_canon_sha256":"3f0a057ff859e968774069b12909d97b2786e1f57687776717abc66d0785e963","abstract_canon_sha256":"e41baa924b7c84a8c651311634366eb397c03503c36920ec0ac0e42a78e490f7"},"schema_version":"1.0"},"canonical_sha256":"8ff6159f9d06b09abe74e55cecfa3ea8500f9575464183abd463ef92835a343a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:00:41.772525Z","signature_b64":"kgKnMkY3VKkfkNYwc0D+dnwuOOmH1B88GEDIfasX27swHOVgK8j/IPNMg4bNJYrc+ciAshQ+7zLTEbMdZBsOAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8ff6159f9d06b09abe74e55cecfa3ea8500f9575464183abd463ef92835a343a","last_reissued_at":"2026-05-20T00:00:41.771762Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:00:41.771762Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.08949","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T00:00:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"C0Ut01aKXvtaK3iSg4mO6iUWodx2iiy1ZlicJHfJh7ajIRBA64RhWf+HsbfK97ml+/7Voz6vCjbMXIG6uxAuDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T00:18:58.829414Z"},"content_sha256":"a1ad140abfcecc4c11c26dc5c14b64d91de271d0085b00339ce16fdd3b2e687c","schema_version":"1.0","event_id":"sha256:a1ad140abfcecc4c11c26dc5c14b64d91de271d0085b00339ce16fdd3b2e687c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:R73BLH45A2YJVPTU4VOOZ6R6VB","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Muon-OGD: Muon-based Spectral Orthogonal Gradient Projection for LLM Continual Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Spectral-norm geometry with orthogonal projections reduces catastrophic forgetting in sequential LLM fine-tuning.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Binghang Lu, Bing Hu, Changhong Mou, Guang Lin, Runyu Zhang, Xiaomin Li, Yuan Tian, Yunhan Zhao, Zheyuan Deng","submitted_at":"2026-05-09T13:42:08Z","abstract_excerpt":"A central challenge in continual learning for large language models (LLMs) is catastrophic forgetting, where adapting to new tasks can substantially degrade performance on previously learned ones. Existing projection-based methods mitigate such interference by restricting parameter updates to subspaces that are orthogonal to directions associated with past tasks. However, these methods are typically formulated under Euclidean parameter geometry, with update magnitudes and projections governed by the Frobenius norm. The recent empirical success of the Muon optimizer, which applies orthogonalize"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Muon-OGD consistently improves over sequential fine-tuning and competitive orthogonal-gradient baselines by applying spectral-norm-aware update geometry with orthogonalized momentum updates that avoid protected directions associated with prior tasks.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That spectral-norm geometry is more suitable than Frobenius norm for matrix-valued LLM parameters, as suggested by the empirical success of the Muon optimizer, and that the resulting constrained optimization can be solved efficiently without compromising the non-interference guarantees.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Muon-OGD introduces a spectral-norm constrained orthogonal projection method solved via dual iterations and Newton-Schulz approximations to improve stability-plasticity trade-off in sequential LLM adaptation.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Spectral-norm geometry with orthogonal projections reduces catastrophic forgetting in sequential LLM fine-tuning.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"a1ef3cb8307caae5188f22b01e24c9b982bdf8814a1314f70ec8b9831d166d04"},"source":{"id":"2605.08949","kind":"arxiv","version":2},"verdict":{"id":"b6f4eabb-b303-442a-a60c-4ffd8db3b48c","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T15:05:02.254800Z","strongest_claim":"Muon-OGD consistently improves over sequential fine-tuning and competitive orthogonal-gradient baselines by applying spectral-norm-aware update geometry with orthogonalized momentum updates that avoid protected directions associated with prior tasks.","one_line_summary":"Muon-OGD introduces a spectral-norm constrained orthogonal projection method solved via dual iterations and Newton-Schulz approximations to improve stability-plasticity trade-off in sequential LLM adaptation.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That spectral-norm geometry is more suitable than Frobenius norm for matrix-valued LLM parameters, as suggested by the empirical success of the Muon optimizer, and that the resulting constrained optimization can be solved efficiently without compromising the non-interference guarantees.","pith_extraction_headline":"Spectral-norm geometry with orthogonal projections reduces catastrophic forgetting in sequential LLM fine-tuning."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.08949/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-19T20:40:48.133391Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T14:01:19.695268Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T10:40:04.544090Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"5b3ea4af26f777c1accaf78771681994fd3bbced91cb786d8eaff71eaeb45e1d"},"references":{"count":48,"sample":[{"doi":"","year":2017,"title":"Overcoming catastrophic forgetting in neural networks.Proceedings of the national academy of sciences, 114(13):3521–3526","work_id":"68b7964f-cd25-47e6-a060-52bbc86a995d","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2013,"title":"An empirical investigation of catastrophic forgetting in gradient-based neural networks.arXiv preprint arXiv:1312.6211","work_id":"2d7055f4-b7d2-4ddb-a9c7-481bd728d360","ref_index":2,"cited_arxiv_id":"1312.6211","is_internal_anchor":true},{"doi":"","year":2017,"title":"Continual learning with deep generative replay.Advances in neural information processing systems, 30","work_id":"4b125fc3-065e-48d4-af0e-bc3c860daf66","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2018,"title":"Lifelong learning with dynamically expandable networks","work_id":"ad0078d3-b62a-4d63-bf25-ce117e735050","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2017,"title":"Learning without forgetting.IEEE transactions on pattern analysis and machine intelligence, 40(12):2935–2947","work_id":"419008ac-740a-4259-ab08-7c02e51962d2","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":48,"snapshot_sha256":"bdf3c10d83ef82b9f3d3b299e4c86748eea28400a6a340fb8b2fb10251fe3def","internal_anchors":8},"formal_canon":{"evidence_count":2,"snapshot_sha256":"06489c60fa81f271710ead4fba8e896751fc3d2c1a7611ed26883eb7a3134eed"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"b6f4eabb-b303-442a-a60c-4ffd8db3b48c"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T00:00:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bqhdeZxPtmKlo3DyaRtCk2x3//RE76BxU6gWJ07Zgqqq8njksAX3c/J28hQNLaTkaqraLNIplkrIdgmxfQVzAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T00:18:58.830365Z"},"content_sha256":"4a486ec4346b9b59f2c87545e2a8a31c57ca17ccf2cc90fa7b41e920ba44db1f","schema_version":"1.0","event_id":"sha256:4a486ec4346b9b59f2c87545e2a8a31c57ca17ccf2cc90fa7b41e920ba44db1f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/R73BLH45A2YJVPTU4VOOZ6R6VB/bundle.json","state_url":"https://pith.science/pith/R73BLH45A2YJVPTU4VOOZ6R6VB/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/R73BLH45A2YJVPTU4VOOZ6R6VB/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":"2026-05-26T00:18:58Z","links":{"resolver":"https://pith.science/pith/R73BLH45A2YJVPTU4VOOZ6R6VB","bundle":"https://pith.science/pith/R73BLH45A2YJVPTU4VOOZ6R6VB/bundle.json","state":"https://pith.science/pith/R73BLH45A2YJVPTU4VOOZ6R6VB/state.json","well_known_bundle":"https://pith.science/.well-known/pith/R73BLH45A2YJVPTU4VOOZ6R6VB/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:R73BLH45A2YJVPTU4VOOZ6R6VB","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":"e41baa924b7c84a8c651311634366eb397c03503c36920ec0ac0e42a78e490f7","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-09T13:42:08Z","title_canon_sha256":"3f0a057ff859e968774069b12909d97b2786e1f57687776717abc66d0785e963"},"schema_version":"1.0","source":{"id":"2605.08949","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.08949","created_at":"2026-05-20T00:00:41Z"},{"alias_kind":"arxiv_version","alias_value":"2605.08949v2","created_at":"2026-05-20T00:00:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.08949","created_at":"2026-05-20T00:00:41Z"},{"alias_kind":"pith_short_12","alias_value":"R73BLH45A2YJ","created_at":"2026-05-20T00:00:41Z"},{"alias_kind":"pith_short_16","alias_value":"R73BLH45A2YJVPTU","created_at":"2026-05-20T00:00:41Z"},{"alias_kind":"pith_short_8","alias_value":"R73BLH45","created_at":"2026-05-20T00:00:41Z"}],"graph_snapshots":[{"event_id":"sha256:4a486ec4346b9b59f2c87545e2a8a31c57ca17ccf2cc90fa7b41e920ba44db1f","target":"graph","created_at":"2026-05-20T00:00:41Z","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":"Muon-OGD consistently improves over sequential fine-tuning and competitive orthogonal-gradient baselines by applying spectral-norm-aware update geometry with orthogonalized momentum updates that avoid protected directions associated with prior tasks."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That spectral-norm geometry is more suitable than Frobenius norm for matrix-valued LLM parameters, as suggested by the empirical success of the Muon optimizer, and that the resulting constrained optimization can be solved efficiently without compromising the non-interference guarantees."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Muon-OGD introduces a spectral-norm constrained orthogonal projection method solved via dual iterations and Newton-Schulz approximations to improve stability-plasticity trade-off in sequential LLM adaptation."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Spectral-norm geometry with orthogonal projections reduces catastrophic forgetting in sequential LLM fine-tuning."}],"snapshot_sha256":"a1ef3cb8307caae5188f22b01e24c9b982bdf8814a1314f70ec8b9831d166d04"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"06489c60fa81f271710ead4fba8e896751fc3d2c1a7611ed26883eb7a3134eed"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-19T20:40:48.133391Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_title_agreement","ran_at":"2026-05-19T14:01:19.695268Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T10:40:04.544090Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2605.08949/integrity.json","findings":[],"snapshot_sha256":"5b3ea4af26f777c1accaf78771681994fd3bbced91cb786d8eaff71eaeb45e1d","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"A central challenge in continual learning for large language models (LLMs) is catastrophic forgetting, where adapting to new tasks can substantially degrade performance on previously learned ones. Existing projection-based methods mitigate such interference by restricting parameter updates to subspaces that are orthogonal to directions associated with past tasks. However, these methods are typically formulated under Euclidean parameter geometry, with update magnitudes and projections governed by the Frobenius norm. The recent empirical success of the Muon optimizer, which applies orthogonalize","authors_text":"Binghang Lu, Bing Hu, Changhong Mou, Guang Lin, Runyu Zhang, Xiaomin Li, Yuan Tian, Yunhan Zhao, Zheyuan Deng","cross_cats":[],"headline":"Spectral-norm geometry with orthogonal projections reduces catastrophic forgetting in sequential LLM fine-tuning.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-09T13:42:08Z","title":"Muon-OGD: Muon-based Spectral Orthogonal Gradient Projection for LLM Continual Learning"},"references":{"count":48,"internal_anchors":8,"resolved_work":48,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Overcoming catastrophic forgetting in neural networks.Proceedings of the national academy of sciences, 114(13):3521–3526","work_id":"68b7964f-cd25-47e6-a060-52bbc86a995d","year":2017},{"cited_arxiv_id":"1312.6211","doi":"","is_internal_anchor":true,"ref_index":2,"title":"An empirical investigation of catastrophic forgetting in gradient-based neural networks.arXiv preprint arXiv:1312.6211","work_id":"2d7055f4-b7d2-4ddb-a9c7-481bd728d360","year":2013},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Continual learning with deep generative replay.Advances in neural information processing systems, 30","work_id":"4b125fc3-065e-48d4-af0e-bc3c860daf66","year":2017},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Lifelong learning with dynamically expandable networks","work_id":"ad0078d3-b62a-4d63-bf25-ce117e735050","year":2018},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Learning without forgetting.IEEE transactions on pattern analysis and machine intelligence, 40(12):2935–2947","work_id":"419008ac-740a-4259-ab08-7c02e51962d2","year":2017}],"snapshot_sha256":"bdf3c10d83ef82b9f3d3b299e4c86748eea28400a6a340fb8b2fb10251fe3def"},"source":{"id":"2605.08949","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-19T15:05:02.254800Z","id":"b6f4eabb-b303-442a-a60c-4ffd8db3b48c","model_set":{"reader":"grok-4.3"},"one_line_summary":"Muon-OGD introduces a spectral-norm constrained orthogonal projection method solved via dual iterations and Newton-Schulz approximations to improve stability-plasticity trade-off in sequential LLM adaptation.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Spectral-norm geometry with orthogonal projections reduces catastrophic forgetting in sequential LLM fine-tuning.","strongest_claim":"Muon-OGD consistently improves over sequential fine-tuning and competitive orthogonal-gradient baselines by applying spectral-norm-aware update geometry with orthogonalized momentum updates that avoid protected directions associated with prior tasks.","weakest_assumption":"That spectral-norm geometry is more suitable than Frobenius norm for matrix-valued LLM parameters, as suggested by the empirical success of the Muon optimizer, and that the resulting constrained optimization can be solved efficiently without compromising the non-interference guarantees."}},"verdict_id":"b6f4eabb-b303-442a-a60c-4ffd8db3b48c"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:a1ad140abfcecc4c11c26dc5c14b64d91de271d0085b00339ce16fdd3b2e687c","target":"record","created_at":"2026-05-20T00:00:41Z","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":"e41baa924b7c84a8c651311634366eb397c03503c36920ec0ac0e42a78e490f7","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-09T13:42:08Z","title_canon_sha256":"3f0a057ff859e968774069b12909d97b2786e1f57687776717abc66d0785e963"},"schema_version":"1.0","source":{"id":"2605.08949","kind":"arxiv","version":2}},"canonical_sha256":"8ff6159f9d06b09abe74e55cecfa3ea8500f9575464183abd463ef92835a343a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8ff6159f9d06b09abe74e55cecfa3ea8500f9575464183abd463ef92835a343a","first_computed_at":"2026-05-20T00:00:41.771762Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:00:41.771762Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"kgKnMkY3VKkfkNYwc0D+dnwuOOmH1B88GEDIfasX27swHOVgK8j/IPNMg4bNJYrc+ciAshQ+7zLTEbMdZBsOAg==","signature_status":"signed_v1","signed_at":"2026-05-20T00:00:41.772525Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.08949","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a1ad140abfcecc4c11c26dc5c14b64d91de271d0085b00339ce16fdd3b2e687c","sha256:4a486ec4346b9b59f2c87545e2a8a31c57ca17ccf2cc90fa7b41e920ba44db1f"],"state_sha256":"677d0296a98613b0189a086eac40af0b454242c0eabcd72fb50e88fba5dd81d6"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZeDWE/sShxVN1hMR9X4Skn2OJxOwWTbFIv7SgVd3yS1fTpWBj7DzqrNUMT9TYwzMoPfnIg2GlSmBrU15pv6QCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-26T00:18:58.834831Z","bundle_sha256":"2283523272ec30cbf3081c21cbc66854c1ca3ef5189681ba8a45ad0515bba0a0"}}