{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2017:M2TSVEMQIXKX2BYL4WBXDNYW6T","short_pith_number":"pith:M2TSVEMQ","canonical_record":{"source":{"id":"1712.02501","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-12-07T06:06:52Z","cross_cats_sorted":["cs.CV","cs.NE"],"title_canon_sha256":"a9bbeaed0dc6b156f9c6cbc481a7385464967eaebc7d5c5b2d1b612395ec9ce5","abstract_canon_sha256":"d3afc22f9b19bc1c1b93964906e50cfd988c33941eb15332dfc751443af15645"},"schema_version":"1.0"},"canonical_sha256":"66a72a919045d57d070be58371b716f4f5d3bb988918a30f8efd202b40ebf9af","source":{"kind":"arxiv","id":"1712.02501","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1712.02501","created_at":"2026-05-18T00:28:01Z"},{"alias_kind":"arxiv_version","alias_value":"1712.02501v2","created_at":"2026-05-18T00:28:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1712.02501","created_at":"2026-05-18T00:28:01Z"},{"alias_kind":"pith_short_12","alias_value":"M2TSVEMQIXKX","created_at":"2026-05-18T12:31:28Z"},{"alias_kind":"pith_short_16","alias_value":"M2TSVEMQIXKX2BYL","created_at":"2026-05-18T12:31:28Z"},{"alias_kind":"pith_short_8","alias_value":"M2TSVEMQ","created_at":"2026-05-18T12:31:28Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2017:M2TSVEMQIXKX2BYL4WBXDNYW6T","target":"record","payload":{"canonical_record":{"source":{"id":"1712.02501","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-12-07T06:06:52Z","cross_cats_sorted":["cs.CV","cs.NE"],"title_canon_sha256":"a9bbeaed0dc6b156f9c6cbc481a7385464967eaebc7d5c5b2d1b612395ec9ce5","abstract_canon_sha256":"d3afc22f9b19bc1c1b93964906e50cfd988c33941eb15332dfc751443af15645"},"schema_version":"1.0"},"canonical_sha256":"66a72a919045d57d070be58371b716f4f5d3bb988918a30f8efd202b40ebf9af","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:28:01.514806Z","signature_b64":"l0TRnhMUvp2y9Ze2Rb0k5did1zsnbTs3TAsv2/JddsU9hpkG1be61TRb8NGXPj7/R3vTZgpmjdkMBTER1I/aCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"66a72a919045d57d070be58371b716f4f5d3bb988918a30f8efd202b40ebf9af","last_reissued_at":"2026-05-18T00:28:01.514192Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:28:01.514192Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1712.02501","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":"2026-05-18T00:28:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ymUQWaHHqFypmiH8qCvcBoObBBZPpECFWMgSSpNeG4EWZr/RqZrxdCIOVd6Cun83RablQImDSnaOxGWQMgFYDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T01:54:36.113145Z"},"content_sha256":"debbb6a9f7a2ddd1b6541227f90b5d65faf0de29b2860ff80c7c55ecd7f7c085","schema_version":"1.0","event_id":"sha256:debbb6a9f7a2ddd1b6541227f90b5d65faf0de29b2860ff80c7c55ecd7f7c085"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2017:M2TSVEMQIXKX2BYL4WBXDNYW6T","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"CNNs are Globally Optimal Given Multi-Layer Support","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.NE"],"primary_cat":"cs.LG","authors_text":"Chen Huang, Chen Kong, Simon Lucey","submitted_at":"2017-12-07T06:06:52Z","abstract_excerpt":"Stochastic Gradient Descent (SGD) is the central workhorse for training modern CNNs. Although giving impressive empirical performance it can be slow to converge. In this paper we explore a novel strategy for training a CNN using an alternation strategy that offers substantial speedups during training. We make the following contributions: (i) replace the ReLU non-linearity within a CNN with positive hard-thresholding, (ii) reinterpret this non-linearity as a binary state vector making the entire CNN linear if the multi-layer support is known, and (iii) demonstrate that under certain conditions "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1712.02501","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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T00:28:01Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"aviR8+gDJJVCKfZ/hAni0ZTcFucvVCKA+tcxa3sbRZRUgKQvqctrVB5j6gOn4w4wvegS1UcMIkLIhAiCKKzVDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T01:54:36.113717Z"},"content_sha256":"357d4e44eaedfc26e7d700e053b640c31dc3b2d81e0e52b0f9c897cb79d019e5","schema_version":"1.0","event_id":"sha256:357d4e44eaedfc26e7d700e053b640c31dc3b2d81e0e52b0f9c897cb79d019e5"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/M2TSVEMQIXKX2BYL4WBXDNYW6T/bundle.json","state_url":"https://pith.science/pith/M2TSVEMQIXKX2BYL4WBXDNYW6T/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/M2TSVEMQIXKX2BYL4WBXDNYW6T/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":"2026-06-05T01:54:36Z","links":{"resolver":"https://pith.science/pith/M2TSVEMQIXKX2BYL4WBXDNYW6T","bundle":"https://pith.science/pith/M2TSVEMQIXKX2BYL4WBXDNYW6T/bundle.json","state":"https://pith.science/pith/M2TSVEMQIXKX2BYL4WBXDNYW6T/state.json","well_known_bundle":"https://pith.science/.well-known/pith/M2TSVEMQIXKX2BYL4WBXDNYW6T/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2017:M2TSVEMQIXKX2BYL4WBXDNYW6T","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":"d3afc22f9b19bc1c1b93964906e50cfd988c33941eb15332dfc751443af15645","cross_cats_sorted":["cs.CV","cs.NE"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-12-07T06:06:52Z","title_canon_sha256":"a9bbeaed0dc6b156f9c6cbc481a7385464967eaebc7d5c5b2d1b612395ec9ce5"},"schema_version":"1.0","source":{"id":"1712.02501","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1712.02501","created_at":"2026-05-18T00:28:01Z"},{"alias_kind":"arxiv_version","alias_value":"1712.02501v2","created_at":"2026-05-18T00:28:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1712.02501","created_at":"2026-05-18T00:28:01Z"},{"alias_kind":"pith_short_12","alias_value":"M2TSVEMQIXKX","created_at":"2026-05-18T12:31:28Z"},{"alias_kind":"pith_short_16","alias_value":"M2TSVEMQIXKX2BYL","created_at":"2026-05-18T12:31:28Z"},{"alias_kind":"pith_short_8","alias_value":"M2TSVEMQ","created_at":"2026-05-18T12:31:28Z"}],"graph_snapshots":[{"event_id":"sha256:357d4e44eaedfc26e7d700e053b640c31dc3b2d81e0e52b0f9c897cb79d019e5","target":"graph","created_at":"2026-05-18T00:28:01Z","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"},"paper":{"abstract_excerpt":"Stochastic Gradient Descent (SGD) is the central workhorse for training modern CNNs. Although giving impressive empirical performance it can be slow to converge. In this paper we explore a novel strategy for training a CNN using an alternation strategy that offers substantial speedups during training. We make the following contributions: (i) replace the ReLU non-linearity within a CNN with positive hard-thresholding, (ii) reinterpret this non-linearity as a binary state vector making the entire CNN linear if the multi-layer support is known, and (iii) demonstrate that under certain conditions ","authors_text":"Chen Huang, Chen Kong, Simon Lucey","cross_cats":["cs.CV","cs.NE"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-12-07T06:06:52Z","title":"CNNs are Globally Optimal Given Multi-Layer Support"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1712.02501","kind":"arxiv","version":2},"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:debbb6a9f7a2ddd1b6541227f90b5d65faf0de29b2860ff80c7c55ecd7f7c085","target":"record","created_at":"2026-05-18T00:28:01Z","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":"d3afc22f9b19bc1c1b93964906e50cfd988c33941eb15332dfc751443af15645","cross_cats_sorted":["cs.CV","cs.NE"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-12-07T06:06:52Z","title_canon_sha256":"a9bbeaed0dc6b156f9c6cbc481a7385464967eaebc7d5c5b2d1b612395ec9ce5"},"schema_version":"1.0","source":{"id":"1712.02501","kind":"arxiv","version":2}},"canonical_sha256":"66a72a919045d57d070be58371b716f4f5d3bb988918a30f8efd202b40ebf9af","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"66a72a919045d57d070be58371b716f4f5d3bb988918a30f8efd202b40ebf9af","first_computed_at":"2026-05-18T00:28:01.514192Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T00:28:01.514192Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"l0TRnhMUvp2y9Ze2Rb0k5did1zsnbTs3TAsv2/JddsU9hpkG1be61TRb8NGXPj7/R3vTZgpmjdkMBTER1I/aCw==","signature_status":"signed_v1","signed_at":"2026-05-18T00:28:01.514806Z","signed_message":"canonical_sha256_bytes"},"source_id":"1712.02501","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:debbb6a9f7a2ddd1b6541227f90b5d65faf0de29b2860ff80c7c55ecd7f7c085","sha256:357d4e44eaedfc26e7d700e053b640c31dc3b2d81e0e52b0f9c897cb79d019e5"],"state_sha256":"1484edae144c51774e596d58ca7de813374a57ea8786ee7392de49dc7e6190ac"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"jbb/M2aEEcRb5eaUipmlOJd7xlBipCtVwuqXSk2upgQJpR95icbJzCKX4FAamJBsFdogXE69gzNeReGSgcuFCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-05T01:54:36.116525Z","bundle_sha256":"652a8615939fe0a8d60e1906ad86a5d21f3930845315d1068ea34bcbc0797262"}}