{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:V6WUGBNWBWXAMXYPFAPDURCXK7","short_pith_number":"pith:V6WUGBNW","schema_version":"1.0","canonical_sha256":"afad4305b60dae065f0f281e3a445757ef98cb47569d532314cfe4d4dfc00632","source":{"kind":"arxiv","id":"2302.01128","version":3},"attestation_state":"computed","paper":{"title":"Mnemosyne: Learning to Train Transformers with Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Avinava Dubey, Deepali Jain, Jie Tan, Krzysztof Marcin Choromanski, Sumeet Singh, Tingnan Zhang, Vikas Sindhwani","submitted_at":"2023-02-02T14:40:28Z","abstract_excerpt":"In this work, we propose a new class of learnable optimizers, called \\textit{Mnemosyne}. It is based on the novel spatio-temporal low-rank implicit attention Transformers that can learn to train entire neural network architectures, including other Transformers, without any task-specific optimizer tuning. We show that Mnemosyne: (a) outperforms popular LSTM optimizers (also with new feature engineering to mitigate catastrophic forgetting of LSTMs), (b) can successfully train Transformers while using simple meta-training strategies that require minimal computational resources, (c) matches accura"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2302.01128","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-02-02T14:40:28Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"21fae680e63046a4b01fb1f7a2f3ee04da7e5f4286f01543149cae4575690614","abstract_canon_sha256":"babad252ce30e3b5d5b42cd5b58d7ef975e4c0a18a9e85643b4016f4a220ef43"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:21:30.410751Z","signature_b64":"Bjp21fnkIHppOFFxzQqBs0GFQmcjDsmWjhB6eA2gLr2ROk5zNZNuMktAfYofFSWgmcv+z3RNV28VUYfQme2XDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"afad4305b60dae065f0f281e3a445757ef98cb47569d532314cfe4d4dfc00632","last_reissued_at":"2026-07-05T06:21:30.410215Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:21:30.410215Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mnemosyne: Learning to Train Transformers with Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Avinava Dubey, Deepali Jain, Jie Tan, Krzysztof Marcin Choromanski, Sumeet Singh, Tingnan Zhang, Vikas Sindhwani","submitted_at":"2023-02-02T14:40:28Z","abstract_excerpt":"In this work, we propose a new class of learnable optimizers, called \\textit{Mnemosyne}. It is based on the novel spatio-temporal low-rank implicit attention Transformers that can learn to train entire neural network architectures, including other Transformers, without any task-specific optimizer tuning. We show that Mnemosyne: (a) outperforms popular LSTM optimizers (also with new feature engineering to mitigate catastrophic forgetting of LSTMs), (b) can successfully train Transformers while using simple meta-training strategies that require minimal computational resources, (c) matches accura"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2302.01128","kind":"arxiv","version":3},"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/2302.01128/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2302.01128","created_at":"2026-07-05T06:21:30.410268+00:00"},{"alias_kind":"arxiv_version","alias_value":"2302.01128v3","created_at":"2026-07-05T06:21:30.410268+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2302.01128","created_at":"2026-07-05T06:21:30.410268+00:00"},{"alias_kind":"pith_short_12","alias_value":"V6WUGBNWBWXA","created_at":"2026-07-05T06:21:30.410268+00:00"},{"alias_kind":"pith_short_16","alias_value":"V6WUGBNWBWXAMXYP","created_at":"2026-07-05T06:21:30.410268+00:00"},{"alias_kind":"pith_short_8","alias_value":"V6WUGBNW","created_at":"2026-07-05T06:21:30.410268+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/V6WUGBNWBWXAMXYPFAPDURCXK7","json":"https://pith.science/pith/V6WUGBNWBWXAMXYPFAPDURCXK7.json","graph_json":"https://pith.science/api/pith-number/V6WUGBNWBWXAMXYPFAPDURCXK7/graph.json","events_json":"https://pith.science/api/pith-number/V6WUGBNWBWXAMXYPFAPDURCXK7/events.json","paper":"https://pith.science/paper/V6WUGBNW"},"agent_actions":{"view_html":"https://pith.science/pith/V6WUGBNWBWXAMXYPFAPDURCXK7","download_json":"https://pith.science/pith/V6WUGBNWBWXAMXYPFAPDURCXK7.json","view_paper":"https://pith.science/paper/V6WUGBNW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2302.01128&json=true","fetch_graph":"https://pith.science/api/pith-number/V6WUGBNWBWXAMXYPFAPDURCXK7/graph.json","fetch_events":"https://pith.science/api/pith-number/V6WUGBNWBWXAMXYPFAPDURCXK7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/V6WUGBNWBWXAMXYPFAPDURCXK7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/V6WUGBNWBWXAMXYPFAPDURCXK7/action/storage_attestation","attest_author":"https://pith.science/pith/V6WUGBNWBWXAMXYPFAPDURCXK7/action/author_attestation","sign_citation":"https://pith.science/pith/V6WUGBNWBWXAMXYPFAPDURCXK7/action/citation_signature","submit_replication":"https://pith.science/pith/V6WUGBNWBWXAMXYPFAPDURCXK7/action/replication_record"}},"created_at":"2026-07-05T06:21:30.410268+00:00","updated_at":"2026-07-05T06:21:30.410268+00:00"}