{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2018:R64ZY3URAAHXIPBE546FZEGY5P","short_pith_number":"pith:R64ZY3UR","canonical_record":{"source":{"id":"1802.05074","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-02-14T12:47:37Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"4279df367d17135f2f25ae7a697becaa7b5bbd56760fed6f91286db38169e464","abstract_canon_sha256":"28fc299ca5531cbb0ae28f5a90dcd8e5b1f2398cc643d22665ed11afdf41692a"},"schema_version":"1.0"},"canonical_sha256":"8fb99c6e91000f743c24ef3c5c90d8ebc802bd9b3be9c17a97b18b0088d9894e","source":{"kind":"arxiv","id":"1802.05074","version":5},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1802.05074","created_at":"2026-05-17T23:59:33Z"},{"alias_kind":"arxiv_version","alias_value":"1802.05074v5","created_at":"2026-05-17T23:59:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1802.05074","created_at":"2026-05-17T23:59:33Z"},{"alias_kind":"pith_short_12","alias_value":"R64ZY3URAAHX","created_at":"2026-05-18T12:32:50Z"},{"alias_kind":"pith_short_16","alias_value":"R64ZY3URAAHXIPBE","created_at":"2026-05-18T12:32:50Z"},{"alias_kind":"pith_short_8","alias_value":"R64ZY3UR","created_at":"2026-05-18T12:32:50Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2018:R64ZY3URAAHXIPBE546FZEGY5P","target":"record","payload":{"canonical_record":{"source":{"id":"1802.05074","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-02-14T12:47:37Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"4279df367d17135f2f25ae7a697becaa7b5bbd56760fed6f91286db38169e464","abstract_canon_sha256":"28fc299ca5531cbb0ae28f5a90dcd8e5b1f2398cc643d22665ed11afdf41692a"},"schema_version":"1.0"},"canonical_sha256":"8fb99c6e91000f743c24ef3c5c90d8ebc802bd9b3be9c17a97b18b0088d9894e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:59:33.022379Z","signature_b64":"M0v4dgj9Xfky38n/32Ct/DWhJY4sa+uIGHMKmi1gv0rLFtiVrtF+l6/4diZH49FTO8XVMdC3B7qsXHKMStlJBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8fb99c6e91000f743c24ef3c5c90d8ebc802bd9b3be9c17a97b18b0088d9894e","last_reissued_at":"2026-05-17T23:59:33.021718Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:59:33.021718Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1802.05074","source_version":5,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:59:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8OteQhlyK768wYteOVHUJARWoOERsmPAvKyxZB562tSRl6OO/50WkNBAwH/z08DvmE2U763yUtN3tOjvLwZ1CQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-02T07:04:30.601502Z"},"content_sha256":"24ca514823982ed8cc257914d718f145ad73e92c825f1b61fb24498ae870882f","schema_version":"1.0","event_id":"sha256:24ca514823982ed8cc257914d718f145ad73e92c825f1b61fb24498ae870882f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2018:R64ZY3URAAHXIPBE546FZEGY5P","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"L4: Practical loss-based stepsize adaptation for deep learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Georg Martius, Michal Rolinek","submitted_at":"2018-02-14T12:47:37Z","abstract_excerpt":"We propose a stepsize adaptation scheme for stochastic gradient descent. It operates directly with the loss function and rescales the gradient in order to make fixed predicted progress on the loss. We demonstrate its capabilities by conclusively improving the performance of Adam and Momentum optimizers. The enhanced optimizers with default hyperparameters consistently outperform their constant stepsize counterparts, even the best ones, without a measurable increase in computational cost. The performance is validated on multiple architectures including dense nets, CNNs, ResNets, and the recurre"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1802.05074","kind":"arxiv","version":5},"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":"2026-05-17T23:59:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"R3573SDCtAZoJva3cLFukjVXb25JbgS3o1lBYtcdmmi71r10qWPEVutgbLkXGOx5Kuzsbh/7Ac8VvHjd1tr5DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-02T07:04:30.601840Z"},"content_sha256":"2f9d4b64169bfa44176b010625d17728872d4140fbd78081842c53a1a30a7ebd","schema_version":"1.0","event_id":"sha256:2f9d4b64169bfa44176b010625d17728872d4140fbd78081842c53a1a30a7ebd"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/R64ZY3URAAHXIPBE546FZEGY5P/bundle.json","state_url":"https://pith.science/pith/R64ZY3URAAHXIPBE546FZEGY5P/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/R64ZY3URAAHXIPBE546FZEGY5P/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":"2026-06-02T07:04:30Z","links":{"resolver":"https://pith.science/pith/R64ZY3URAAHXIPBE546FZEGY5P","bundle":"https://pith.science/pith/R64ZY3URAAHXIPBE546FZEGY5P/bundle.json","state":"https://pith.science/pith/R64ZY3URAAHXIPBE546FZEGY5P/state.json","well_known_bundle":"https://pith.science/.well-known/pith/R64ZY3URAAHXIPBE546FZEGY5P/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2018:R64ZY3URAAHXIPBE546FZEGY5P","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":"28fc299ca5531cbb0ae28f5a90dcd8e5b1f2398cc643d22665ed11afdf41692a","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-02-14T12:47:37Z","title_canon_sha256":"4279df367d17135f2f25ae7a697becaa7b5bbd56760fed6f91286db38169e464"},"schema_version":"1.0","source":{"id":"1802.05074","kind":"arxiv","version":5}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1802.05074","created_at":"2026-05-17T23:59:33Z"},{"alias_kind":"arxiv_version","alias_value":"1802.05074v5","created_at":"2026-05-17T23:59:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1802.05074","created_at":"2026-05-17T23:59:33Z"},{"alias_kind":"pith_short_12","alias_value":"R64ZY3URAAHX","created_at":"2026-05-18T12:32:50Z"},{"alias_kind":"pith_short_16","alias_value":"R64ZY3URAAHXIPBE","created_at":"2026-05-18T12:32:50Z"},{"alias_kind":"pith_short_8","alias_value":"R64ZY3UR","created_at":"2026-05-18T12:32:50Z"}],"graph_snapshots":[{"event_id":"sha256:2f9d4b64169bfa44176b010625d17728872d4140fbd78081842c53a1a30a7ebd","target":"graph","created_at":"2026-05-17T23:59:33Z","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":"We propose a stepsize adaptation scheme for stochastic gradient descent. It operates directly with the loss function and rescales the gradient in order to make fixed predicted progress on the loss. We demonstrate its capabilities by conclusively improving the performance of Adam and Momentum optimizers. The enhanced optimizers with default hyperparameters consistently outperform their constant stepsize counterparts, even the best ones, without a measurable increase in computational cost. The performance is validated on multiple architectures including dense nets, CNNs, ResNets, and the recurre","authors_text":"Georg Martius, Michal Rolinek","cross_cats":["stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-02-14T12:47:37Z","title":"L4: Practical loss-based stepsize adaptation for deep learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1802.05074","kind":"arxiv","version":5},"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":"sha256:24ca514823982ed8cc257914d718f145ad73e92c825f1b61fb24498ae870882f","target":"record","created_at":"2026-05-17T23:59:33Z","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":"28fc299ca5531cbb0ae28f5a90dcd8e5b1f2398cc643d22665ed11afdf41692a","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-02-14T12:47:37Z","title_canon_sha256":"4279df367d17135f2f25ae7a697becaa7b5bbd56760fed6f91286db38169e464"},"schema_version":"1.0","source":{"id":"1802.05074","kind":"arxiv","version":5}},"canonical_sha256":"8fb99c6e91000f743c24ef3c5c90d8ebc802bd9b3be9c17a97b18b0088d9894e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8fb99c6e91000f743c24ef3c5c90d8ebc802bd9b3be9c17a97b18b0088d9894e","first_computed_at":"2026-05-17T23:59:33.021718Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:59:33.021718Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"M0v4dgj9Xfky38n/32Ct/DWhJY4sa+uIGHMKmi1gv0rLFtiVrtF+l6/4diZH49FTO8XVMdC3B7qsXHKMStlJBQ==","signature_status":"signed_v1","signed_at":"2026-05-17T23:59:33.022379Z","signed_message":"canonical_sha256_bytes"},"source_id":"1802.05074","source_kind":"arxiv","source_version":5}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:24ca514823982ed8cc257914d718f145ad73e92c825f1b61fb24498ae870882f","sha256:2f9d4b64169bfa44176b010625d17728872d4140fbd78081842c53a1a30a7ebd"],"state_sha256":"a7690c76beb44f4c8decef004791f59d432025d94f2a19f50c63aa12688fb70e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"45idY4CEmKoTIM9uZfPZRqRuCjJxzTNaC7ON6iohKhIFxVsqD+x6qayt6qTWsh81HXA3+YMMPPwGcC36nOP0BA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-02T07:04:30.603650Z","bundle_sha256":"a5e923ceccdde80532e57167f8a869ecccbe9c9c9f099a4bf4b121045f1ab03f"}}