{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:TKDOH5VEMQIPF6XGFLPWEXLJ74","short_pith_number":"pith:TKDOH5VE","canonical_record":{"source":{"id":"2112.15383","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2021-12-31T10:49:55Z","cross_cats_sorted":["cs.LG","physics.data-an"],"title_canon_sha256":"0d1d5a90ac5df761c473141633247bc7d76883399cd2b6ae10a5ec541dee9ec7","abstract_canon_sha256":"e66e15c823b68b148041597258040fa3919fdb0a9ea5346abafe7e7b69bcc977"},"schema_version":"1.0"},"canonical_sha256":"9a86e3f6a46410f2fae62adf625d69ff233a892d6b42082f341a080dfaefc53b","source":{"kind":"arxiv","id":"2112.15383","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2112.15383","created_at":"2026-07-05T05:00:14Z"},{"alias_kind":"arxiv_version","alias_value":"2112.15383v3","created_at":"2026-07-05T05:00:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2112.15383","created_at":"2026-07-05T05:00:14Z"},{"alias_kind":"pith_short_12","alias_value":"TKDOH5VEMQIP","created_at":"2026-07-05T05:00:14Z"},{"alias_kind":"pith_short_16","alias_value":"TKDOH5VEMQIPF6XG","created_at":"2026-07-05T05:00:14Z"},{"alias_kind":"pith_short_8","alias_value":"TKDOH5VE","created_at":"2026-07-05T05:00:14Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:TKDOH5VEMQIPF6XGFLPWEXLJ74","target":"record","payload":{"canonical_record":{"source":{"id":"2112.15383","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2021-12-31T10:49:55Z","cross_cats_sorted":["cs.LG","physics.data-an"],"title_canon_sha256":"0d1d5a90ac5df761c473141633247bc7d76883399cd2b6ae10a5ec541dee9ec7","abstract_canon_sha256":"e66e15c823b68b148041597258040fa3919fdb0a9ea5346abafe7e7b69bcc977"},"schema_version":"1.0"},"canonical_sha256":"9a86e3f6a46410f2fae62adf625d69ff233a892d6b42082f341a080dfaefc53b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:00:14.783632Z","signature_b64":"ZE2UPBya4TdBb5rZfcSl3QcVuC1mNPP5SC3N+0mSrKYFmkVR3VOfgh4kXUMLMOTEAFdPKgL/5fo7QSjBKad2Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9a86e3f6a46410f2fae62adf625d69ff233a892d6b42082f341a080dfaefc53b","last_reissued_at":"2026-07-05T05:00:14.783194Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:00:14.783194Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2112.15383","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T05:00:14Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ohncAIQNllOmSNWTzGqSFikxv5S+/7XafxvJiqqU/HUHnY9Ksp52C72HD4llb0j3vIasjyKiHSrm25SZmonDDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T09:02:16.618153Z"},"content_sha256":"5df108a7614478be9b0c2cea90c3ee245f0e08d2aa031d0b9d902a1f5a99e0e5","schema_version":"1.0","event_id":"sha256:5df108a7614478be9b0c2cea90c3ee245f0e08d2aa031d0b9d902a1f5a99e0e5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:TKDOH5VEMQIPF6XGFLPWEXLJ74","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Separation of Scales and a Thermodynamic Description of Feature Learning in Some CNNs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","physics.data-an"],"primary_cat":"stat.ML","authors_text":"Gadi Naveh, Inbar Seroussi, Zohar Ringel","submitted_at":"2021-12-31T10:49:55Z","abstract_excerpt":"Deep neural networks (DNNs) are powerful tools for compressing and distilling information. Their scale and complexity, often involving billions of inter-dependent parameters, render direct microscopic analysis difficult. Under such circumstances, a common strategy is to identify slow variables that average the erratic behavior of the fast microscopic variables. Here, we identify a similar separation of scales occurring in fully trained finitely over-parameterized deep convolutional neural networks (CNNs) and fully connected networks (FCNs). Specifically, we show that DNN layers couple only thr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2112.15383","kind":"arxiv","version":3},"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/2112.15383/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T05:00:14Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0q86e8JpPWNghU7+gXv5iWw7ftTzvSR01O2z+vzxbTrWo73MAHBCEcutnt6IHqGWOc6XzlZMvT4diNo3FPJiAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T09:02:16.619084Z"},"content_sha256":"ae40605ec21721826afa9fcbd5be1ee783fddfd081aed182f18990f102d92ee4","schema_version":"1.0","event_id":"sha256:ae40605ec21721826afa9fcbd5be1ee783fddfd081aed182f18990f102d92ee4"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/TKDOH5VEMQIPF6XGFLPWEXLJ74/bundle.json","state_url":"https://pith.science/pith/TKDOH5VEMQIPF6XGFLPWEXLJ74/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/TKDOH5VEMQIPF6XGFLPWEXLJ74/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":"2026-08-09T09:02:16Z","links":{"resolver":"https://pith.science/pith/TKDOH5VEMQIPF6XGFLPWEXLJ74","bundle":"https://pith.science/pith/TKDOH5VEMQIPF6XGFLPWEXLJ74/bundle.json","state":"https://pith.science/pith/TKDOH5VEMQIPF6XGFLPWEXLJ74/state.json","well_known_bundle":"https://pith.science/.well-known/pith/TKDOH5VEMQIPF6XGFLPWEXLJ74/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:TKDOH5VEMQIPF6XGFLPWEXLJ74","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":"e66e15c823b68b148041597258040fa3919fdb0a9ea5346abafe7e7b69bcc977","cross_cats_sorted":["cs.LG","physics.data-an"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2021-12-31T10:49:55Z","title_canon_sha256":"0d1d5a90ac5df761c473141633247bc7d76883399cd2b6ae10a5ec541dee9ec7"},"schema_version":"1.0","source":{"id":"2112.15383","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2112.15383","created_at":"2026-07-05T05:00:14Z"},{"alias_kind":"arxiv_version","alias_value":"2112.15383v3","created_at":"2026-07-05T05:00:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2112.15383","created_at":"2026-07-05T05:00:14Z"},{"alias_kind":"pith_short_12","alias_value":"TKDOH5VEMQIP","created_at":"2026-07-05T05:00:14Z"},{"alias_kind":"pith_short_16","alias_value":"TKDOH5VEMQIPF6XG","created_at":"2026-07-05T05:00:14Z"},{"alias_kind":"pith_short_8","alias_value":"TKDOH5VE","created_at":"2026-07-05T05:00:14Z"}],"graph_snapshots":[{"event_id":"sha256:ae40605ec21721826afa9fcbd5be1ee783fddfd081aed182f18990f102d92ee4","target":"graph","created_at":"2026-07-05T05:00:14Z","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":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2112.15383/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Deep neural networks (DNNs) are powerful tools for compressing and distilling information. Their scale and complexity, often involving billions of inter-dependent parameters, render direct microscopic analysis difficult. Under such circumstances, a common strategy is to identify slow variables that average the erratic behavior of the fast microscopic variables. Here, we identify a similar separation of scales occurring in fully trained finitely over-parameterized deep convolutional neural networks (CNNs) and fully connected networks (FCNs). Specifically, we show that DNN layers couple only thr","authors_text":"Gadi Naveh, Inbar Seroussi, Zohar Ringel","cross_cats":["cs.LG","physics.data-an"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2021-12-31T10:49:55Z","title":"Separation of Scales and a Thermodynamic Description of Feature Learning in Some CNNs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2112.15383","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:5df108a7614478be9b0c2cea90c3ee245f0e08d2aa031d0b9d902a1f5a99e0e5","target":"record","created_at":"2026-07-05T05:00:14Z","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":"e66e15c823b68b148041597258040fa3919fdb0a9ea5346abafe7e7b69bcc977","cross_cats_sorted":["cs.LG","physics.data-an"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2021-12-31T10:49:55Z","title_canon_sha256":"0d1d5a90ac5df761c473141633247bc7d76883399cd2b6ae10a5ec541dee9ec7"},"schema_version":"1.0","source":{"id":"2112.15383","kind":"arxiv","version":3}},"canonical_sha256":"9a86e3f6a46410f2fae62adf625d69ff233a892d6b42082f341a080dfaefc53b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9a86e3f6a46410f2fae62adf625d69ff233a892d6b42082f341a080dfaefc53b","first_computed_at":"2026-07-05T05:00:14.783194Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T05:00:14.783194Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ZE2UPBya4TdBb5rZfcSl3QcVuC1mNPP5SC3N+0mSrKYFmkVR3VOfgh4kXUMLMOTEAFdPKgL/5fo7QSjBKad2Cg==","signature_status":"signed_v1","signed_at":"2026-07-05T05:00:14.783632Z","signed_message":"canonical_sha256_bytes"},"source_id":"2112.15383","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5df108a7614478be9b0c2cea90c3ee245f0e08d2aa031d0b9d902a1f5a99e0e5","sha256:ae40605ec21721826afa9fcbd5be1ee783fddfd081aed182f18990f102d92ee4"],"state_sha256":"7eaa0666f31b9d6cf2c6019bd45b868bf10c6dfd3689b9c833d0b20e23c4d56e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"H5S9X8enyxztKkM863TuhOzjt04oeNCZntiK06tgm8I0uNRwRK8L4EKrd+07rpS/yA9pJ0rop5clYr+AZtBUAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T09:02:16.626672Z","bundle_sha256":"c6f2d422ae2f4011d33cd9ac675000bc06e0f16a7dd1301831294fe24602cc3e"}}