{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2020:DLSADBD2LEEUIRAENK4OWUN7KO","short_pith_number":"pith:DLSADBD2","canonical_record":{"source":{"id":"2011.02408","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2020-11-04T17:04:01Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"dd810bd1fe38820d6b178e68ce8dc95c986c356ec3cab3dfbc9accd20c377f1f","abstract_canon_sha256":"f6ff4e37c518fc3e1322bb7f87f81992ab9a879b0fb0d5597b9120744097ddcb"},"schema_version":"1.0"},"canonical_sha256":"1ae401847a59094444046ab8eb51bf539bbbf85bfe8f0382e9ad86c17e53137f","source":{"kind":"arxiv","id":"2011.02408","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2011.02408","created_at":"2026-07-05T06:53:27Z"},{"alias_kind":"arxiv_version","alias_value":"2011.02408v2","created_at":"2026-07-05T06:53:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2011.02408","created_at":"2026-07-05T06:53:27Z"},{"alias_kind":"pith_short_12","alias_value":"DLSADBD2LEEU","created_at":"2026-07-05T06:53:27Z"},{"alias_kind":"pith_short_16","alias_value":"DLSADBD2LEEUIRAE","created_at":"2026-07-05T06:53:27Z"},{"alias_kind":"pith_short_8","alias_value":"DLSADBD2","created_at":"2026-07-05T06:53:27Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2020:DLSADBD2LEEUIRAENK4OWUN7KO","target":"record","payload":{"canonical_record":{"source":{"id":"2011.02408","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2020-11-04T17:04:01Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"dd810bd1fe38820d6b178e68ce8dc95c986c356ec3cab3dfbc9accd20c377f1f","abstract_canon_sha256":"f6ff4e37c518fc3e1322bb7f87f81992ab9a879b0fb0d5597b9120744097ddcb"},"schema_version":"1.0"},"canonical_sha256":"1ae401847a59094444046ab8eb51bf539bbbf85bfe8f0382e9ad86c17e53137f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:53:27.867240Z","signature_b64":"zkDWVWGYHJODrLKW23JrmG075RM5MPNBxjjFHBENUMiShPI6jNpOlgKTThkvt/ZKvfAzNqHkQgAaCcIe3t97Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1ae401847a59094444046ab8eb51bf539bbbf85bfe8f0382e9ad86c17e53137f","last_reissued_at":"2026-07-05T06:53:27.866750Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:53:27.866750Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2011.02408","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":"2026-07-05T06:53:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EGqgg3wtrrfAnXDCbBeYsyMmij/7nUsirL/rcGBRZKAzFtU3im6Y9LQi9nc+mETIiQ89turBxtpGhv7KSIvDDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-08T16:23:18.065485Z"},"content_sha256":"a5cd5e63e79def0fa9cf02edc96c4cbb913d5a73bef56c22fd4c14649210d899","schema_version":"1.0","event_id":"sha256:a5cd5e63e79def0fa9cf02edc96c4cbb913d5a73bef56c22fd4c14649210d899"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2020:DLSADBD2LEEUIRAENK4OWUN7KO","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Which Minimizer Does My Neural Network Converge To?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"David Reeb, Ingo Steinwart, Manuel Nonnenmacher","submitted_at":"2020-11-04T17:04:01Z","abstract_excerpt":"The loss surface of an overparameterized neural network (NN) possesses many global minima of zero training error. We explain how common variants of the standard NN training procedure change the minimizer obtained. First, we make explicit how the size of the initialization of a strongly overparameterized NN affects the minimizer and can deteriorate its final test performance. We propose a strategy to limit this effect. Then, we demonstrate that for adaptive optimization such as AdaGrad, the obtained minimizer generally differs from the gradient descent (GD) minimizer. This adaptive minimizer is"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2011.02408","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2011.02408/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":"2026-07-05T06:53:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kHkn9H4vGMhDOpnvMJ3jHoqkaxnx4iRSssbdJoNpKnYhdc2hqk30v7c6KPhqL6O66mj40XDu4YLZq9LFyTs9AQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-08T16:23:18.065903Z"},"content_sha256":"9d8b9bc08410a8a49c7b85191c61f4281fb91a05679380f8eefab198e4fb9833","schema_version":"1.0","event_id":"sha256:9d8b9bc08410a8a49c7b85191c61f4281fb91a05679380f8eefab198e4fb9833"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DLSADBD2LEEUIRAENK4OWUN7KO/bundle.json","state_url":"https://pith.science/pith/DLSADBD2LEEUIRAENK4OWUN7KO/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DLSADBD2LEEUIRAENK4OWUN7KO/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":"2026-07-08T16:23:18Z","links":{"resolver":"https://pith.science/pith/DLSADBD2LEEUIRAENK4OWUN7KO","bundle":"https://pith.science/pith/DLSADBD2LEEUIRAENK4OWUN7KO/bundle.json","state":"https://pith.science/pith/DLSADBD2LEEUIRAENK4OWUN7KO/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DLSADBD2LEEUIRAENK4OWUN7KO/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:DLSADBD2LEEUIRAENK4OWUN7KO","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":"f6ff4e37c518fc3e1322bb7f87f81992ab9a879b0fb0d5597b9120744097ddcb","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2020-11-04T17:04:01Z","title_canon_sha256":"dd810bd1fe38820d6b178e68ce8dc95c986c356ec3cab3dfbc9accd20c377f1f"},"schema_version":"1.0","source":{"id":"2011.02408","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2011.02408","created_at":"2026-07-05T06:53:27Z"},{"alias_kind":"arxiv_version","alias_value":"2011.02408v2","created_at":"2026-07-05T06:53:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2011.02408","created_at":"2026-07-05T06:53:27Z"},{"alias_kind":"pith_short_12","alias_value":"DLSADBD2LEEU","created_at":"2026-07-05T06:53:27Z"},{"alias_kind":"pith_short_16","alias_value":"DLSADBD2LEEUIRAE","created_at":"2026-07-05T06:53:27Z"},{"alias_kind":"pith_short_8","alias_value":"DLSADBD2","created_at":"2026-07-05T06:53:27Z"}],"graph_snapshots":[{"event_id":"sha256:9d8b9bc08410a8a49c7b85191c61f4281fb91a05679380f8eefab198e4fb9833","target":"graph","created_at":"2026-07-05T06:53:27Z","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/2011.02408/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The loss surface of an overparameterized neural network (NN) possesses many global minima of zero training error. We explain how common variants of the standard NN training procedure change the minimizer obtained. First, we make explicit how the size of the initialization of a strongly overparameterized NN affects the minimizer and can deteriorate its final test performance. We propose a strategy to limit this effect. Then, we demonstrate that for adaptive optimization such as AdaGrad, the obtained minimizer generally differs from the gradient descent (GD) minimizer. This adaptive minimizer is","authors_text":"David Reeb, Ingo Steinwart, Manuel Nonnenmacher","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2020-11-04T17:04:01Z","title":"Which Minimizer Does My Neural Network Converge To?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2011.02408","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:a5cd5e63e79def0fa9cf02edc96c4cbb913d5a73bef56c22fd4c14649210d899","target":"record","created_at":"2026-07-05T06:53:27Z","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":"f6ff4e37c518fc3e1322bb7f87f81992ab9a879b0fb0d5597b9120744097ddcb","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2020-11-04T17:04:01Z","title_canon_sha256":"dd810bd1fe38820d6b178e68ce8dc95c986c356ec3cab3dfbc9accd20c377f1f"},"schema_version":"1.0","source":{"id":"2011.02408","kind":"arxiv","version":2}},"canonical_sha256":"1ae401847a59094444046ab8eb51bf539bbbf85bfe8f0382e9ad86c17e53137f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"1ae401847a59094444046ab8eb51bf539bbbf85bfe8f0382e9ad86c17e53137f","first_computed_at":"2026-07-05T06:53:27.866750Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T06:53:27.866750Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"zkDWVWGYHJODrLKW23JrmG075RM5MPNBxjjFHBENUMiShPI6jNpOlgKTThkvt/ZKvfAzNqHkQgAaCcIe3t97Cw==","signature_status":"signed_v1","signed_at":"2026-07-05T06:53:27.867240Z","signed_message":"canonical_sha256_bytes"},"source_id":"2011.02408","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a5cd5e63e79def0fa9cf02edc96c4cbb913d5a73bef56c22fd4c14649210d899","sha256:9d8b9bc08410a8a49c7b85191c61f4281fb91a05679380f8eefab198e4fb9833"],"state_sha256":"dd54f15c77ddd9c395de896e3acd91a78d008c41de57360df7edfadc06570c98"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QGVjdd5Ou6BnutT9B99vMxMDnvrlQ1NQNrwRBN0teGPP/8sM5LXgHbLQIJf5J9x4n+Y8atxw9oCthuDxoewPBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-08T16:23:18.068065Z","bundle_sha256":"8f77a90b6c09d6e60f919c0b548f8ffa5cbb30a6573d9d3c8032d4dd5c97226a"}}