{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2020:D5ZXEAJ7RRJ7EVGRSGDHQ3N522","short_pith_number":"pith:D5ZXEAJ7","canonical_record":{"source":{"id":"2003.02389","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-03-05T00:53:18Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"30022641ae97e73b62c5bb44ef1e60f574cc0573d974b8c28a271568e1c4136d","abstract_canon_sha256":"1c711fbfc1149f351cda3cf2966ccc25601ab49305ff6ec382cb4bdf87850d0c"},"schema_version":"1.0"},"canonical_sha256":"1f7372013f8c53f254d19186786dbdd69f741b27d31a309b1ab212888898a6a6","source":{"kind":"arxiv","id":"2003.02389","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2003.02389","created_at":"2026-07-05T00:45:56Z"},{"alias_kind":"arxiv_version","alias_value":"2003.02389v1","created_at":"2026-07-05T00:45:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2003.02389","created_at":"2026-07-05T00:45:56Z"},{"alias_kind":"pith_short_12","alias_value":"D5ZXEAJ7RRJ7","created_at":"2026-07-05T00:45:56Z"},{"alias_kind":"pith_short_16","alias_value":"D5ZXEAJ7RRJ7EVGR","created_at":"2026-07-05T00:45:56Z"},{"alias_kind":"pith_short_8","alias_value":"D5ZXEAJ7","created_at":"2026-07-05T00:45:56Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2020:D5ZXEAJ7RRJ7EVGRSGDHQ3N522","target":"record","payload":{"canonical_record":{"source":{"id":"2003.02389","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-03-05T00:53:18Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"30022641ae97e73b62c5bb44ef1e60f574cc0573d974b8c28a271568e1c4136d","abstract_canon_sha256":"1c711fbfc1149f351cda3cf2966ccc25601ab49305ff6ec382cb4bdf87850d0c"},"schema_version":"1.0"},"canonical_sha256":"1f7372013f8c53f254d19186786dbdd69f741b27d31a309b1ab212888898a6a6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:45:56.025257Z","signature_b64":"szN9j0gFuAb35ECxQSFZRSNDVIqWVryfyLw67ZHsmPkQx8KITcmE1ksGuLGd0OlskiNsYVh7MiPb64uk3DJGAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1f7372013f8c53f254d19186786dbdd69f741b27d31a309b1ab212888898a6a6","last_reissued_at":"2026-07-05T00:45:56.024787Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:45:56.024787Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2003.02389","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":"2026-07-05T00:45:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SmEIeKZXX3XQoHQmHHMJ6qtoY1+7yuPdEwU8k+ZOUr718EQ8ChiRKDSgt9io+h7mYmCj4ISue3UvvL467kXbCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T08:27:41.841413Z"},"content_sha256":"678f1cefe1be39d88c8a63bf9d3fedcb377088a3bbb288f393692fa24817a66c","schema_version":"1.0","event_id":"sha256:678f1cefe1be39d88c8a63bf9d3fedcb377088a3bbb288f393692fa24817a66c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2020:D5ZXEAJ7RRJ7EVGRSGDHQ3N522","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Comparing Rewinding and Fine-tuning in Neural Network Pruning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Alex Renda, Jonathan Frankle, Michael Carbin","submitted_at":"2020-03-05T00:53:18Z","abstract_excerpt":"Many neural network pruning algorithms proceed in three steps: train the network to completion, remove unwanted structure to compress the network, and retrain the remaining structure to recover lost accuracy. The standard retraining technique, fine-tuning, trains the unpruned weights from their final trained values using a small fixed learning rate. In this paper, we compare fine-tuning to alternative retraining techniques. Weight rewinding (as proposed by Frankle et al., (2019)), rewinds unpruned weights to their values from earlier in training and retrains them from there using the original "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2003.02389","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/2003.02389/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":"2026-07-05T00:45:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"dMxBd/+iJrQWa47kL9v9r33XdC6OVAB9cE86Z4Pcyjp7/zwKrtpGoaT8DtA8RhnUCQ0gmQYeR9+PqVmUwAeaAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T08:27:41.841783Z"},"content_sha256":"23e83ec0dc68dabcd71e437692ce7ea2d5ac82cf24a03f4120ee0c8f492a7727","schema_version":"1.0","event_id":"sha256:23e83ec0dc68dabcd71e437692ce7ea2d5ac82cf24a03f4120ee0c8f492a7727"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/D5ZXEAJ7RRJ7EVGRSGDHQ3N522/bundle.json","state_url":"https://pith.science/pith/D5ZXEAJ7RRJ7EVGRSGDHQ3N522/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/D5ZXEAJ7RRJ7EVGRSGDHQ3N522/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":"2026-07-07T08:27:41Z","links":{"resolver":"https://pith.science/pith/D5ZXEAJ7RRJ7EVGRSGDHQ3N522","bundle":"https://pith.science/pith/D5ZXEAJ7RRJ7EVGRSGDHQ3N522/bundle.json","state":"https://pith.science/pith/D5ZXEAJ7RRJ7EVGRSGDHQ3N522/state.json","well_known_bundle":"https://pith.science/.well-known/pith/D5ZXEAJ7RRJ7EVGRSGDHQ3N522/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:D5ZXEAJ7RRJ7EVGRSGDHQ3N522","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":"1c711fbfc1149f351cda3cf2966ccc25601ab49305ff6ec382cb4bdf87850d0c","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-03-05T00:53:18Z","title_canon_sha256":"30022641ae97e73b62c5bb44ef1e60f574cc0573d974b8c28a271568e1c4136d"},"schema_version":"1.0","source":{"id":"2003.02389","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2003.02389","created_at":"2026-07-05T00:45:56Z"},{"alias_kind":"arxiv_version","alias_value":"2003.02389v1","created_at":"2026-07-05T00:45:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2003.02389","created_at":"2026-07-05T00:45:56Z"},{"alias_kind":"pith_short_12","alias_value":"D5ZXEAJ7RRJ7","created_at":"2026-07-05T00:45:56Z"},{"alias_kind":"pith_short_16","alias_value":"D5ZXEAJ7RRJ7EVGR","created_at":"2026-07-05T00:45:56Z"},{"alias_kind":"pith_short_8","alias_value":"D5ZXEAJ7","created_at":"2026-07-05T00:45:56Z"}],"graph_snapshots":[{"event_id":"sha256:23e83ec0dc68dabcd71e437692ce7ea2d5ac82cf24a03f4120ee0c8f492a7727","target":"graph","created_at":"2026-07-05T00:45:56Z","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/2003.02389/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Many neural network pruning algorithms proceed in three steps: train the network to completion, remove unwanted structure to compress the network, and retrain the remaining structure to recover lost accuracy. The standard retraining technique, fine-tuning, trains the unpruned weights from their final trained values using a small fixed learning rate. In this paper, we compare fine-tuning to alternative retraining techniques. Weight rewinding (as proposed by Frankle et al., (2019)), rewinds unpruned weights to their values from earlier in training and retrains them from there using the original ","authors_text":"Alex Renda, Jonathan Frankle, Michael Carbin","cross_cats":["stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-03-05T00:53:18Z","title":"Comparing Rewinding and Fine-tuning in Neural Network Pruning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2003.02389","kind":"arxiv","version":1},"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":"sha256:678f1cefe1be39d88c8a63bf9d3fedcb377088a3bbb288f393692fa24817a66c","target":"record","created_at":"2026-07-05T00:45:56Z","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":"1c711fbfc1149f351cda3cf2966ccc25601ab49305ff6ec382cb4bdf87850d0c","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-03-05T00:53:18Z","title_canon_sha256":"30022641ae97e73b62c5bb44ef1e60f574cc0573d974b8c28a271568e1c4136d"},"schema_version":"1.0","source":{"id":"2003.02389","kind":"arxiv","version":1}},"canonical_sha256":"1f7372013f8c53f254d19186786dbdd69f741b27d31a309b1ab212888898a6a6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"1f7372013f8c53f254d19186786dbdd69f741b27d31a309b1ab212888898a6a6","first_computed_at":"2026-07-05T00:45:56.024787Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T00:45:56.024787Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"szN9j0gFuAb35ECxQSFZRSNDVIqWVryfyLw67ZHsmPkQx8KITcmE1ksGuLGd0OlskiNsYVh7MiPb64uk3DJGAA==","signature_status":"signed_v1","signed_at":"2026-07-05T00:45:56.025257Z","signed_message":"canonical_sha256_bytes"},"source_id":"2003.02389","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:678f1cefe1be39d88c8a63bf9d3fedcb377088a3bbb288f393692fa24817a66c","sha256:23e83ec0dc68dabcd71e437692ce7ea2d5ac82cf24a03f4120ee0c8f492a7727"],"state_sha256":"8ba1712a117dd2e08975fd0fc790d93f83b36cf5d7b571ba4caef7cd683173f1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/F6rx4QevsyB5497tcsUbqdhYnTalPDrTWnrgGQdhDiWlKFTlpfZKUy4jx2mWR099AUZCJ0rXFUKfYRdkRc/Cg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T08:27:41.843728Z","bundle_sha256":"334e7c9abe429271065b8721267a59dcda5c41d7e9c143fcc2355f2cc3556a90"}}