{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:5TMGYSGQU7GVMKHTO2PFMLFQ43","short_pith_number":"pith:5TMGYSGQ","schema_version":"1.0","canonical_sha256":"ecd86c48d0a7cd5628f3769e562cb0e6eeb93c5c653565865b49a9a612ebc451","source":{"kind":"arxiv","id":"2606.09762","version":1},"attestation_state":"computed","paper":{"title":"Preserving Plasticity in Continual Learning via Dynamical Isometry","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Andries Rosseau, Ann Now\\'e, Robert M\\\"uller","submitted_at":"2026-06-08T17:24:15Z","abstract_excerpt":"Continual training of deep neural networks under non-stationarity often leads to a progressive loss of plasticity, eventually limiting further learning. We relate plasticity to the empirical Neural Tangent Kernel, and identify dynamical isometry (the condition that layer-wise Jacobian singular values remain close to one) as a key mechanism for preserving plasticity in continual learning. We revisit a class of networks that are almost-everywhere isometric while remaining universal Lipschitz function approximators, demonstrating that near-dynamical isometry is compatible with expressive nonlinea"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2606.09762","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-08T17:24:15Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1b61e43aab77b72aa4a49cb68216f17789df62483d194e9811a376c4f9eabe43","abstract_canon_sha256":"71bce03075ebae2f58a3d72090509e0820fc6e354144dff5a5becacf6d59bb48"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T02:09:08.277397Z","signature_b64":"HB0qFIgKAyPh3hsrKCHp9vfP3F4hJas3AfVYUbDnzV9X3/ExOcBiLaNV+O4iJmZ98cPZJhb+dh66iOk02H4/Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ecd86c48d0a7cd5628f3769e562cb0e6eeb93c5c653565865b49a9a612ebc451","last_reissued_at":"2026-06-09T02:09:08.276514Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T02:09:08.276514Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Preserving Plasticity in Continual Learning via Dynamical Isometry","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Andries Rosseau, Ann Now\\'e, Robert M\\\"uller","submitted_at":"2026-06-08T17:24:15Z","abstract_excerpt":"Continual training of deep neural networks under non-stationarity often leads to a progressive loss of plasticity, eventually limiting further learning. We relate plasticity to the empirical Neural Tangent Kernel, and identify dynamical isometry (the condition that layer-wise Jacobian singular values remain close to one) as a key mechanism for preserving plasticity in continual learning. We revisit a class of networks that are almost-everywhere isometric while remaining universal Lipschitz function approximators, demonstrating that near-dynamical isometry is compatible with expressive nonlinea"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.09762","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2606.09762/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2606.09762","created_at":"2026-06-09T02:09:08.276645+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.09762v1","created_at":"2026-06-09T02:09:08.276645+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.09762","created_at":"2026-06-09T02:09:08.276645+00:00"},{"alias_kind":"pith_short_12","alias_value":"5TMGYSGQU7GV","created_at":"2026-06-09T02:09:08.276645+00:00"},{"alias_kind":"pith_short_16","alias_value":"5TMGYSGQU7GVMKHT","created_at":"2026-06-09T02:09:08.276645+00:00"},{"alias_kind":"pith_short_8","alias_value":"5TMGYSGQ","created_at":"2026-06-09T02:09:08.276645+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5TMGYSGQU7GVMKHTO2PFMLFQ43","json":"https://pith.science/pith/5TMGYSGQU7GVMKHTO2PFMLFQ43.json","graph_json":"https://pith.science/api/pith-number/5TMGYSGQU7GVMKHTO2PFMLFQ43/graph.json","events_json":"https://pith.science/api/pith-number/5TMGYSGQU7GVMKHTO2PFMLFQ43/events.json","paper":"https://pith.science/paper/5TMGYSGQ"},"agent_actions":{"view_html":"https://pith.science/pith/5TMGYSGQU7GVMKHTO2PFMLFQ43","download_json":"https://pith.science/pith/5TMGYSGQU7GVMKHTO2PFMLFQ43.json","view_paper":"https://pith.science/paper/5TMGYSGQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.09762&json=true","fetch_graph":"https://pith.science/api/pith-number/5TMGYSGQU7GVMKHTO2PFMLFQ43/graph.json","fetch_events":"https://pith.science/api/pith-number/5TMGYSGQU7GVMKHTO2PFMLFQ43/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5TMGYSGQU7GVMKHTO2PFMLFQ43/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5TMGYSGQU7GVMKHTO2PFMLFQ43/action/storage_attestation","attest_author":"https://pith.science/pith/5TMGYSGQU7GVMKHTO2PFMLFQ43/action/author_attestation","sign_citation":"https://pith.science/pith/5TMGYSGQU7GVMKHTO2PFMLFQ43/action/citation_signature","submit_replication":"https://pith.science/pith/5TMGYSGQU7GVMKHTO2PFMLFQ43/action/replication_record"}},"created_at":"2026-06-09T02:09:08.276645+00:00","updated_at":"2026-06-09T02:09:08.276645+00:00"}