{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:N7VNTZ53FADWCOTYZJZFZOAAKP","short_pith_number":"pith:N7VNTZ53","canonical_record":{"source":{"id":"2605.15435","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T21:27:38Z","cross_cats_sorted":["cs.NE"],"title_canon_sha256":"a74fc4bd2da7991ad7aa4f56c09c6bec5feebd376c938e07b6ff38d3e0a9bc8b","abstract_canon_sha256":"574a18fc44f44579d08d94a288d86ab448c1272d31bd7848eb8c4e06ad9bcc99"},"schema_version":"1.0"},"canonical_sha256":"6fead9e7bb2807613a78ca725cb80053e3db709d4c624fc75281b696d33503f2","source":{"kind":"arxiv","id":"2605.15435","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.15435","created_at":"2026-05-20T00:00:58Z"},{"alias_kind":"arxiv_version","alias_value":"2605.15435v1","created_at":"2026-05-20T00:00:58Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.15435","created_at":"2026-05-20T00:00:58Z"},{"alias_kind":"pith_short_12","alias_value":"N7VNTZ53FADW","created_at":"2026-05-20T00:00:58Z"},{"alias_kind":"pith_short_16","alias_value":"N7VNTZ53FADWCOTY","created_at":"2026-05-20T00:00:58Z"},{"alias_kind":"pith_short_8","alias_value":"N7VNTZ53","created_at":"2026-05-20T00:00:58Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:N7VNTZ53FADWCOTYZJZFZOAAKP","target":"record","payload":{"canonical_record":{"source":{"id":"2605.15435","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T21:27:38Z","cross_cats_sorted":["cs.NE"],"title_canon_sha256":"a74fc4bd2da7991ad7aa4f56c09c6bec5feebd376c938e07b6ff38d3e0a9bc8b","abstract_canon_sha256":"574a18fc44f44579d08d94a288d86ab448c1272d31bd7848eb8c4e06ad9bcc99"},"schema_version":"1.0"},"canonical_sha256":"6fead9e7bb2807613a78ca725cb80053e3db709d4c624fc75281b696d33503f2","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:00:58.476988Z","signature_b64":"6btXN3NyQiHxI3eno6h0b1u9yaT6oXxYwelVRG8mhqAsnab7MsxZVyAOKG+E1Pp48mvJGIBWfCtzriy/8IuOAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6fead9e7bb2807613a78ca725cb80053e3db709d4c624fc75281b696d33503f2","last_reissued_at":"2026-05-20T00:00:58.476189Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:00:58.476189Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.15435","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":"2026-05-20T00:00:58Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5UoeeCP4dEZGwowX5jGUeR8QJjudZyht/UxTZiOaKzSfFXtMqslacf1/XgI4dujHRgasjrtKzRh3paQbgb7NBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-22T17:58:00.334829Z"},"content_sha256":"f682159b8ac430402a2ec7a86c7fa9f88decea10e494ae0dfa40788ab33875f5","schema_version":"1.0","event_id":"sha256:f682159b8ac430402a2ec7a86c7fa9f88decea10e494ae0dfa40788ab33875f5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:N7VNTZ53FADWCOTYZJZFZOAAKP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"On the Stability of Growth in Structural Plasticity","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Newborn network units participate in the forward pass but receive weaker gradients than units present from the start.","cross_cats":["cs.NE"],"primary_cat":"cs.LG","authors_text":"Lute Lillo, Nick Cheney","submitted_at":"2026-05-14T21:27:38Z","abstract_excerpt":"Standard deep-learning pipelines usually choose the network architecture before training and keep it fixed throughout optimization. In contrast, a model can also be adapted by editing its structure during training, for example by pruning existing hidden-neuron units or growing new ones. Although growth is appealing for adaptive and continual systems, we show that it is not simply the inverse of pruning. Pruning selects among units that have participated in training from the start, whereas growth inserts new units into an already specialized optimization trajectory. We isolate this insertion pr"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"newborn units are often forward-active but backward-starved: they participate in the forward computation, yet receive much weaker gradient signal than incumbent units.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The optimization trajectory has already become specialized by the time new units are inserted, so that their gradient signal is systematically weaker than that of units present from the start.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Newborn units in growing neural networks are forward-active but backward-starved, receiving weaker gradients than existing units and creating integration challenges that make growth less reliable than pruning in complex tasks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Newborn network units participate in the forward pass but receive weaker gradients than units present from the start.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"673a48fdb8c5d326d0d7f857a981f405de3fd4c8a12f0d47279918fffb523ce5"},"source":{"id":"2605.15435","kind":"arxiv","version":1},"verdict":{"id":"087c5718-7e55-439e-a847-d1e77a188b46","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T15:34:07.989745Z","strongest_claim":"newborn units are often forward-active but backward-starved: they participate in the forward computation, yet receive much weaker gradient signal than incumbent units.","one_line_summary":"Newborn units in growing neural networks are forward-active but backward-starved, receiving weaker gradients than existing units and creating integration challenges that make growth less reliable than pruning in complex tasks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The optimization trajectory has already become specialized by the time new units are inserted, so that their gradient signal is systematically weaker than that of units present from the start.","pith_extraction_headline":"Newborn network units participate in the forward pass but receive weaker gradients than units present from the start."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.15435/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T16:01:17.979298Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"cited_work_retraction","ran_at":"2026-05-19T15:54:27.976475Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"citation_quote_validity","ran_at":"2026-05-19T15:50:29.344268Z","status":"completed","version":"0.1.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T15:40:36.907332Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T14:21:54.126306Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T13:33:22.691155Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"8794ab52cf756246b5bdd71c08639d158715b386e22cd4ecae45a27b6491cc09"},"references":{"count":55,"sample":[{"doi":"","year":2023,"title":"Asadi, K., Fakoor, R., and Sabach, S. (2023). Resetting the optimizer in deep rl: An empirical study.Advances in Neural Information Processing Systems, 36:72284–72324","work_id":"566d5324-2e6c-4e2d-85f2-da1857c3d537","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Nested learning: The illusion of deep learning architectures.arXiv preprint arXiv:2512.24695, 2025a","work_id":"faa7da49-a9f6-4714-a8c8-5692f267f0b1","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2017,"title":"Deep Rewiring: Training very sparse deep networks","work_id":"7b32e667-baf1-464d-89bb-5104c770ba6d","ref_index":3,"cited_arxiv_id":"1711.05136","is_internal_anchor":true},{"doi":"","year":2021,"title":"Cai, Z., Sener, O., and Koltun, V. (2021). Online continual learning with natural distribution shifts: An empirical study with visual data. InProceedings of the IEEE/CVF international conference on co","work_id":"72a3f968-f35d-4ca3-9808-1b60a69ff9f4","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2016,"title":"Chen, T., Goodfellow, I., and Shlens, J. (2016). Net2net: Accelerating learning via knowledge transfer. InICLR","work_id":"3aa79721-e2eb-4b54-af7c-d9400136c843","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":55,"snapshot_sha256":"44f181dfaf7539919260cef0f6b3cd53cc8a7924a87b51e9e43906c37c5438af","internal_anchors":13},"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":"087c5718-7e55-439e-a847-d1e77a188b46"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T00:00:58Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZjXk8DwbQ9nHCKg8Csve15KnMd3JztXcmqBBr3v8/qMJY4NWNThFh5/fY5cde+rVpF8BjOrglQtIkLHwrEGQCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-22T17:58:00.335569Z"},"content_sha256":"69f766676dce9f7bd47ffc43a426560cd950ca00d5ecf76f6e3b54552b82980b","schema_version":"1.0","event_id":"sha256:69f766676dce9f7bd47ffc43a426560cd950ca00d5ecf76f6e3b54552b82980b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/N7VNTZ53FADWCOTYZJZFZOAAKP/bundle.json","state_url":"https://pith.science/pith/N7VNTZ53FADWCOTYZJZFZOAAKP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/N7VNTZ53FADWCOTYZJZFZOAAKP/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":"2026-05-22T17:58:00Z","links":{"resolver":"https://pith.science/pith/N7VNTZ53FADWCOTYZJZFZOAAKP","bundle":"https://pith.science/pith/N7VNTZ53FADWCOTYZJZFZOAAKP/bundle.json","state":"https://pith.science/pith/N7VNTZ53FADWCOTYZJZFZOAAKP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/N7VNTZ53FADWCOTYZJZFZOAAKP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:N7VNTZ53FADWCOTYZJZFZOAAKP","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":"574a18fc44f44579d08d94a288d86ab448c1272d31bd7848eb8c4e06ad9bcc99","cross_cats_sorted":["cs.NE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T21:27:38Z","title_canon_sha256":"a74fc4bd2da7991ad7aa4f56c09c6bec5feebd376c938e07b6ff38d3e0a9bc8b"},"schema_version":"1.0","source":{"id":"2605.15435","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.15435","created_at":"2026-05-20T00:00:58Z"},{"alias_kind":"arxiv_version","alias_value":"2605.15435v1","created_at":"2026-05-20T00:00:58Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.15435","created_at":"2026-05-20T00:00:58Z"},{"alias_kind":"pith_short_12","alias_value":"N7VNTZ53FADW","created_at":"2026-05-20T00:00:58Z"},{"alias_kind":"pith_short_16","alias_value":"N7VNTZ53FADWCOTY","created_at":"2026-05-20T00:00:58Z"},{"alias_kind":"pith_short_8","alias_value":"N7VNTZ53","created_at":"2026-05-20T00:00:58Z"}],"graph_snapshots":[{"event_id":"sha256:69f766676dce9f7bd47ffc43a426560cd950ca00d5ecf76f6e3b54552b82980b","target":"graph","created_at":"2026-05-20T00:00:58Z","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":"newborn units are often forward-active but backward-starved: they participate in the forward computation, yet receive much weaker gradient signal than incumbent units."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The optimization trajectory has already become specialized by the time new units are inserted, so that their gradient signal is systematically weaker than that of units present from the start."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Newborn units in growing neural networks are forward-active but backward-starved, receiving weaker gradients than existing units and creating integration challenges that make growth less reliable than pruning in complex tasks."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Newborn network units participate in the forward pass but receive weaker gradients than units present from the start."}],"snapshot_sha256":"673a48fdb8c5d326d0d7f857a981f405de3fd4c8a12f0d47279918fffb523ce5"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"doi_title_agreement","ran_at":"2026-05-19T16:01:17.979298Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"cited_work_retraction","ran_at":"2026-05-19T15:54:27.976475Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"citation_quote_validity","ran_at":"2026-05-19T15:50:29.344268Z","status":"completed","version":"0.1.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T15:40:36.907332Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"claim_evidence","ran_at":"2026-05-19T14:21:54.126306Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-19T13:33:22.691155Z","status":"skipped","version":"1.0.0"}],"endpoint":"/pith/2605.15435/integrity.json","findings":[],"snapshot_sha256":"8794ab52cf756246b5bdd71c08639d158715b386e22cd4ecae45a27b6491cc09","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Standard deep-learning pipelines usually choose the network architecture before training and keep it fixed throughout optimization. In contrast, a model can also be adapted by editing its structure during training, for example by pruning existing hidden-neuron units or growing new ones. Although growth is appealing for adaptive and continual systems, we show that it is not simply the inverse of pruning. Pruning selects among units that have participated in training from the start, whereas growth inserts new units into an already specialized optimization trajectory. We isolate this insertion pr","authors_text":"Lute Lillo, Nick Cheney","cross_cats":["cs.NE"],"headline":"Newborn network units participate in the forward pass but receive weaker gradients than units present from the start.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T21:27:38Z","title":"On the Stability of Growth in Structural Plasticity"},"references":{"count":55,"internal_anchors":13,"resolved_work":55,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Asadi, K., Fakoor, R., and Sabach, S. (2023). Resetting the optimizer in deep rl: An empirical study.Advances in Neural Information Processing Systems, 36:72284–72324","work_id":"566d5324-2e6c-4e2d-85f2-da1857c3d537","year":2023},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Nested learning: The illusion of deep learning architectures.arXiv preprint arXiv:2512.24695, 2025a","work_id":"faa7da49-a9f6-4714-a8c8-5692f267f0b1","year":2025},{"cited_arxiv_id":"1711.05136","doi":"","is_internal_anchor":true,"ref_index":3,"title":"Deep Rewiring: Training very sparse deep networks","work_id":"7b32e667-baf1-464d-89bb-5104c770ba6d","year":2017},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Cai, Z., Sener, O., and Koltun, V. (2021). Online continual learning with natural distribution shifts: An empirical study with visual data. InProceedings of the IEEE/CVF international conference on co","work_id":"72a3f968-f35d-4ca3-9808-1b60a69ff9f4","year":2021},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Chen, T., Goodfellow, I., and Shlens, J. (2016). Net2net: Accelerating learning via knowledge transfer. InICLR","work_id":"3aa79721-e2eb-4b54-af7c-d9400136c843","year":2016}],"snapshot_sha256":"44f181dfaf7539919260cef0f6b3cd53cc8a7924a87b51e9e43906c37c5438af"},"source":{"id":"2605.15435","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-19T15:34:07.989745Z","id":"087c5718-7e55-439e-a847-d1e77a188b46","model_set":{"reader":"grok-4.3"},"one_line_summary":"Newborn units in growing neural networks are forward-active but backward-starved, receiving weaker gradients than existing units and creating integration challenges that make growth less reliable than pruning in complex tasks.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Newborn network units participate in the forward pass but receive weaker gradients than units present from the start.","strongest_claim":"newborn units are often forward-active but backward-starved: they participate in the forward computation, yet receive much weaker gradient signal than incumbent units.","weakest_assumption":"The optimization trajectory has already become specialized by the time new units are inserted, so that their gradient signal is systematically weaker than that of units present from the start."}},"verdict_id":"087c5718-7e55-439e-a847-d1e77a188b46"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:f682159b8ac430402a2ec7a86c7fa9f88decea10e494ae0dfa40788ab33875f5","target":"record","created_at":"2026-05-20T00:00:58Z","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":"574a18fc44f44579d08d94a288d86ab448c1272d31bd7848eb8c4e06ad9bcc99","cross_cats_sorted":["cs.NE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T21:27:38Z","title_canon_sha256":"a74fc4bd2da7991ad7aa4f56c09c6bec5feebd376c938e07b6ff38d3e0a9bc8b"},"schema_version":"1.0","source":{"id":"2605.15435","kind":"arxiv","version":1}},"canonical_sha256":"6fead9e7bb2807613a78ca725cb80053e3db709d4c624fc75281b696d33503f2","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6fead9e7bb2807613a78ca725cb80053e3db709d4c624fc75281b696d33503f2","first_computed_at":"2026-05-20T00:00:58.476189Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:00:58.476189Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"6btXN3NyQiHxI3eno6h0b1u9yaT6oXxYwelVRG8mhqAsnab7MsxZVyAOKG+E1Pp48mvJGIBWfCtzriy/8IuOAQ==","signature_status":"signed_v1","signed_at":"2026-05-20T00:00:58.476988Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.15435","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f682159b8ac430402a2ec7a86c7fa9f88decea10e494ae0dfa40788ab33875f5","sha256:69f766676dce9f7bd47ffc43a426560cd950ca00d5ecf76f6e3b54552b82980b"],"state_sha256":"ca73de984200402c6fb952992719d800718505f74bb4ca8da63c0ca52d6e1542"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"P1KbtVRq9k2UngThcZILalHfJCW3dAPKd2hNrV3IcGI9Il9d3y1o/79+N0pzrafvcGECQmppvuppPl/h5tlEAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-22T17:58:00.338439Z","bundle_sha256":"1d6e28d7fb275f4a0d9f2f144e801204a3cf6d1236aed0910bd30e4b78620a29"}}