{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:M3PJQ35KX43OGJAVKNV672LJDE","short_pith_number":"pith:M3PJQ35K","schema_version":"1.0","canonical_sha256":"66de986faabf36e32415536befe9691911cb0fe24d8f0ea1db8069be4187b442","source":{"kind":"arxiv","id":"1706.05806","version":2},"attestation_state":"computed","paper":{"title":"SVCCA: Singular Vector Canonical Correlation Analysis for Deep Learning Dynamics and Interpretability","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Jascha Sohl-Dickstein, Jason Yosinski, Justin Gilmer, Maithra Raghu","submitted_at":"2017-06-19T07:09:20Z","abstract_excerpt":"We propose a new technique, Singular Vector Canonical Correlation Analysis (SVCCA), a tool for quickly comparing two representations in a way that is both invariant to affine transform (allowing comparison between different layers and networks) and fast to compute (allowing more comparisons to be calculated than with previous methods). We deploy this tool to measure the intrinsic dimensionality of layers, showing in some cases needless over-parameterization; to probe learning dynamics throughout training, finding that networks converge to final representations from the bottom up; to show where"},"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":"1706.05806","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2017-06-19T07:09:20Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"cf866b37cb98f3bfe07553cae54eef2574f3d579b0baf0048d51b025d010c2e1","abstract_canon_sha256":"903ef2b930ad24545e6bcba5cd0e68a178a94e8f8cd9281b0048b970f5e73a0d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:31:02.741126Z","signature_b64":"wqeC0vrmT0cEl2deGBfKBfkIOOolie5KGrOAIMYPA5DdgK0bFzqYuqyHIZUyaiM97Vip1kUsKSel9MVs/J9QAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"66de986faabf36e32415536befe9691911cb0fe24d8f0ea1db8069be4187b442","last_reissued_at":"2026-05-18T00:31:02.740432Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:31:02.740432Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SVCCA: Singular Vector Canonical Correlation Analysis for Deep Learning Dynamics and Interpretability","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Jascha Sohl-Dickstein, Jason Yosinski, Justin Gilmer, Maithra Raghu","submitted_at":"2017-06-19T07:09:20Z","abstract_excerpt":"We propose a new technique, Singular Vector Canonical Correlation Analysis (SVCCA), a tool for quickly comparing two representations in a way that is both invariant to affine transform (allowing comparison between different layers and networks) and fast to compute (allowing more comparisons to be calculated than with previous methods). We deploy this tool to measure the intrinsic dimensionality of layers, showing in some cases needless over-parameterization; to probe learning dynamics throughout training, finding that networks converge to final representations from the bottom up; to show where"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1706.05806","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"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":"1706.05806","created_at":"2026-05-18T00:31:02.740544+00:00"},{"alias_kind":"arxiv_version","alias_value":"1706.05806v2","created_at":"2026-05-18T00:31:02.740544+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1706.05806","created_at":"2026-05-18T00:31:02.740544+00:00"},{"alias_kind":"pith_short_12","alias_value":"M3PJQ35KX43O","created_at":"2026-05-18T12:31:28.150371+00:00"},{"alias_kind":"pith_short_16","alias_value":"M3PJQ35KX43OGJAV","created_at":"2026-05-18T12:31:28.150371+00:00"},{"alias_kind":"pith_short_8","alias_value":"M3PJQ35K","created_at":"2026-05-18T12:31:28.150371+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":5,"sample":[{"citing_arxiv_id":"2607.00924","citing_title":"Graph-Native Reinforcement Learning Enables Traceable Scientific Hypothesis Generation through Conceptual Recombination","ref_index":55,"is_internal_anchor":true},{"citing_arxiv_id":"2605.15901","citing_title":"From Layers to Networks: Comparing Neural Representations via Diffusion Geometry","ref_index":10,"is_internal_anchor":true},{"citing_arxiv_id":"2605.17704","citing_title":"Toy Combinatorial Interpretability Models Reveal Lottery Tickets in Early Feature Space","ref_index":68,"is_internal_anchor":true},{"citing_arxiv_id":"2605.15183","citing_title":"When Are Two Networks the Same? Tensor Similarity for Mechanistic Interpretability","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2604.13082","citing_title":"The Long Delay to Arithmetic Generalization: When Learned Representations Outrun Behavior","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"1610.01644","citing_title":"Understanding intermediate layers using linear classifier probes","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/M3PJQ35KX43OGJAVKNV672LJDE","json":"https://pith.science/pith/M3PJQ35KX43OGJAVKNV672LJDE.json","graph_json":"https://pith.science/api/pith-number/M3PJQ35KX43OGJAVKNV672LJDE/graph.json","events_json":"https://pith.science/api/pith-number/M3PJQ35KX43OGJAVKNV672LJDE/events.json","paper":"https://pith.science/paper/M3PJQ35K"},"agent_actions":{"view_html":"https://pith.science/pith/M3PJQ35KX43OGJAVKNV672LJDE","download_json":"https://pith.science/pith/M3PJQ35KX43OGJAVKNV672LJDE.json","view_paper":"https://pith.science/paper/M3PJQ35K","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1706.05806&json=true","fetch_graph":"https://pith.science/api/pith-number/M3PJQ35KX43OGJAVKNV672LJDE/graph.json","fetch_events":"https://pith.science/api/pith-number/M3PJQ35KX43OGJAVKNV672LJDE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/M3PJQ35KX43OGJAVKNV672LJDE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/M3PJQ35KX43OGJAVKNV672LJDE/action/storage_attestation","attest_author":"https://pith.science/pith/M3PJQ35KX43OGJAVKNV672LJDE/action/author_attestation","sign_citation":"https://pith.science/pith/M3PJQ35KX43OGJAVKNV672LJDE/action/citation_signature","submit_replication":"https://pith.science/pith/M3PJQ35KX43OGJAVKNV672LJDE/action/replication_record"}},"created_at":"2026-05-18T00:31:02.740544+00:00","updated_at":"2026-05-18T00:31:02.740544+00:00"}