{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:FMPUFZW2HWX7DSEISETUJIHR4A","short_pith_number":"pith:FMPUFZW2","canonical_record":{"source":{"id":"2402.03496","kind":"arxiv","version":10},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-05T20:15:19Z","cross_cats_sorted":["math.OC"],"title_canon_sha256":"97acf2c38f60bdf73bf4e325168c13423487230d9586afd4dc703943091a7389","abstract_canon_sha256":"6f09acc25493b3eefe51bd6bdcc278212ed771e16d7349229371b98e57b60d84"},"schema_version":"1.0"},"canonical_sha256":"2b1f42e6da3daff1c888912744a0f1e00d579ff08f6339ac04f2d7dc3bcec0ac","source":{"kind":"arxiv","id":"2402.03496","version":10},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.03496","created_at":"2026-07-05T09:16:12Z"},{"alias_kind":"arxiv_version","alias_value":"2402.03496v10","created_at":"2026-07-05T09:16:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.03496","created_at":"2026-07-05T09:16:12Z"},{"alias_kind":"pith_short_12","alias_value":"FMPUFZW2HWX7","created_at":"2026-07-05T09:16:12Z"},{"alias_kind":"pith_short_16","alias_value":"FMPUFZW2HWX7DSEI","created_at":"2026-07-05T09:16:12Z"},{"alias_kind":"pith_short_8","alias_value":"FMPUFZW2","created_at":"2026-07-05T09:16:12Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:FMPUFZW2HWX7DSEISETUJIHR4A","target":"record","payload":{"canonical_record":{"source":{"id":"2402.03496","kind":"arxiv","version":10},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-05T20:15:19Z","cross_cats_sorted":["math.OC"],"title_canon_sha256":"97acf2c38f60bdf73bf4e325168c13423487230d9586afd4dc703943091a7389","abstract_canon_sha256":"6f09acc25493b3eefe51bd6bdcc278212ed771e16d7349229371b98e57b60d84"},"schema_version":"1.0"},"canonical_sha256":"2b1f42e6da3daff1c888912744a0f1e00d579ff08f6339ac04f2d7dc3bcec0ac","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:16:12.393182Z","signature_b64":"She9SQtYlAYFrs6RyM62PAaCzfKFSSfOii6BOCRS3jxSfC9VWB3M6DKep4VODNRayxLsjCy9EzWhuo7SmVjuDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2b1f42e6da3daff1c888912744a0f1e00d579ff08f6339ac04f2d7dc3bcec0ac","last_reissued_at":"2026-07-05T09:16:12.392689Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:16:12.392689Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2402.03496","source_version":10,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:16:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Uf73CU1o8kES+52Nq3ZPhvLW5ybgRhO9zajc6qIM7/0fLBgIZn91CxjHXcXCD9CCtQMJi6CT1GXIw38opRSiDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T13:30:38.494632Z"},"content_sha256":"ce97584dbd0d9eec90c71915ec4b931a990cfc55237790b565928aafb8eab1d3","schema_version":"1.0","event_id":"sha256:ce97584dbd0d9eec90c71915ec4b931a990cfc55237790b565928aafb8eab1d3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:FMPUFZW2HWX7DSEISETUJIHR4A","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Can We Remove the Square-Root in Adaptive Gradient Methods? A Second-Order Perspective","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["math.OC"],"primary_cat":"cs.LG","authors_text":"Alireza Makhzani, Felix Dangel, Juhan Bae, Richard E. Turner, Runa Eschenhagen, Wu Lin","submitted_at":"2024-02-05T20:15:19Z","abstract_excerpt":"Adaptive gradient optimizers like Adam(W) are the default training algorithms for many deep learning architectures, such as transformers. Their diagonal preconditioner is based on the gradient outer product which is incorporated into the parameter update via a square root. While these methods are often motivated as approximate second-order methods, the square root represents a fundamental difference. In this work, we investigate how the behavior of adaptive methods changes when we remove the root, i.e., strengthen their second-order motivation. Surprisingly, we find that such square-root-free "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.03496","kind":"arxiv","version":10},"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/2402.03496/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":"2026-07-05T09:16:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6pqCwwryuuKVPGYQPxpPEzYR1Tt3hUUQAiIqiflNkYTYDNWAGyWwiFAuEpvE+7DeIKL9MaSRq+Q2w4VqgpaSBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T13:30:38.495567Z"},"content_sha256":"8ba98ae328480d7b02339bd0ab08ae630e6aab7c10382130f6b859c4ab381b0b","schema_version":"1.0","event_id":"sha256:8ba98ae328480d7b02339bd0ab08ae630e6aab7c10382130f6b859c4ab381b0b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FMPUFZW2HWX7DSEISETUJIHR4A/bundle.json","state_url":"https://pith.science/pith/FMPUFZW2HWX7DSEISETUJIHR4A/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FMPUFZW2HWX7DSEISETUJIHR4A/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":"2026-08-05T13:30:38Z","links":{"resolver":"https://pith.science/pith/FMPUFZW2HWX7DSEISETUJIHR4A","bundle":"https://pith.science/pith/FMPUFZW2HWX7DSEISETUJIHR4A/bundle.json","state":"https://pith.science/pith/FMPUFZW2HWX7DSEISETUJIHR4A/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FMPUFZW2HWX7DSEISETUJIHR4A/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:FMPUFZW2HWX7DSEISETUJIHR4A","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":"6f09acc25493b3eefe51bd6bdcc278212ed771e16d7349229371b98e57b60d84","cross_cats_sorted":["math.OC"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-05T20:15:19Z","title_canon_sha256":"97acf2c38f60bdf73bf4e325168c13423487230d9586afd4dc703943091a7389"},"schema_version":"1.0","source":{"id":"2402.03496","kind":"arxiv","version":10}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.03496","created_at":"2026-07-05T09:16:12Z"},{"alias_kind":"arxiv_version","alias_value":"2402.03496v10","created_at":"2026-07-05T09:16:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.03496","created_at":"2026-07-05T09:16:12Z"},{"alias_kind":"pith_short_12","alias_value":"FMPUFZW2HWX7","created_at":"2026-07-05T09:16:12Z"},{"alias_kind":"pith_short_16","alias_value":"FMPUFZW2HWX7DSEI","created_at":"2026-07-05T09:16:12Z"},{"alias_kind":"pith_short_8","alias_value":"FMPUFZW2","created_at":"2026-07-05T09:16:12Z"}],"graph_snapshots":[{"event_id":"sha256:8ba98ae328480d7b02339bd0ab08ae630e6aab7c10382130f6b859c4ab381b0b","target":"graph","created_at":"2026-07-05T09:16:12Z","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/2402.03496/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Adaptive gradient optimizers like Adam(W) are the default training algorithms for many deep learning architectures, such as transformers. Their diagonal preconditioner is based on the gradient outer product which is incorporated into the parameter update via a square root. While these methods are often motivated as approximate second-order methods, the square root represents a fundamental difference. In this work, we investigate how the behavior of adaptive methods changes when we remove the root, i.e., strengthen their second-order motivation. Surprisingly, we find that such square-root-free ","authors_text":"Alireza Makhzani, Felix Dangel, Juhan Bae, Richard E. Turner, Runa Eschenhagen, Wu Lin","cross_cats":["math.OC"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-05T20:15:19Z","title":"Can We Remove the Square-Root in Adaptive Gradient Methods? A Second-Order Perspective"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.03496","kind":"arxiv","version":10},"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:ce97584dbd0d9eec90c71915ec4b931a990cfc55237790b565928aafb8eab1d3","target":"record","created_at":"2026-07-05T09:16:12Z","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":"6f09acc25493b3eefe51bd6bdcc278212ed771e16d7349229371b98e57b60d84","cross_cats_sorted":["math.OC"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-05T20:15:19Z","title_canon_sha256":"97acf2c38f60bdf73bf4e325168c13423487230d9586afd4dc703943091a7389"},"schema_version":"1.0","source":{"id":"2402.03496","kind":"arxiv","version":10}},"canonical_sha256":"2b1f42e6da3daff1c888912744a0f1e00d579ff08f6339ac04f2d7dc3bcec0ac","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2b1f42e6da3daff1c888912744a0f1e00d579ff08f6339ac04f2d7dc3bcec0ac","first_computed_at":"2026-07-05T09:16:12.392689Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:16:12.392689Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"She9SQtYlAYFrs6RyM62PAaCzfKFSSfOii6BOCRS3jxSfC9VWB3M6DKep4VODNRayxLsjCy9EzWhuo7SmVjuDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T09:16:12.393182Z","signed_message":"canonical_sha256_bytes"},"source_id":"2402.03496","source_kind":"arxiv","source_version":10}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ce97584dbd0d9eec90c71915ec4b931a990cfc55237790b565928aafb8eab1d3","sha256:8ba98ae328480d7b02339bd0ab08ae630e6aab7c10382130f6b859c4ab381b0b"],"state_sha256":"a2ff3f49624dcbf40544ad60d3f4511fe4cbc57eca3076d81b30a05f005e5369"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"khdNqRZiyNKIUwJ2n0DoEMJtiLjHOAXnoiygFBWBlqfRtZYNdx4KOL3ozE7r1vPtJ0tsU5Xc/90IlRWZDle6DQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T13:30:38.510527Z","bundle_sha256":"27c91de2eedc57c2ead4757d938cc7add2de0750c3b19a815d09aa539241b35f"}}