{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:4TN3LGYDBSIKGR3HAPSRXUORYE","short_pith_number":"pith:4TN3LGYD","schema_version":"1.0","canonical_sha256":"e4dbb59b030c90a3476703e51bd1d1c116524ebdd154479f31ab1c90fe829049","source":{"kind":"arxiv","id":"2303.15725","version":2},"attestation_state":"computed","paper":{"title":"Solving Regularized Exp, Cosh and Sinh Regression Problems","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Tianyi Zhou, Zhao Song, Zhihang Li","submitted_at":"2023-03-28T04:26:51Z","abstract_excerpt":"In modern machine learning, attention computation is a fundamental task for training large language models such as Transformer, GPT-4 and ChatGPT. In this work, we study exponential regression problem which is inspired by the softmax/exp unit in the attention mechanism in large language models. The standard exponential regression is non-convex. We study the regularization version of exponential regression problem which is a convex problem. We use approximate newton method to solve in input sparsity time.\n  Formally, in this problem, one is given matrix $A \\in \\mathbb{R}^{n \\times d}$, $b \\in \\"},"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":"2303.15725","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2023-03-28T04:26:51Z","cross_cats_sorted":[],"title_canon_sha256":"15c62f4b68f295ac07cba3b2e3a4837b702f21a0dc37da3c0717a87c912d2a8f","abstract_canon_sha256":"ccbe02e3ac560d512841bc5b1f8814c7a4cffb664e2c61399bbee3598a3de831"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:09:09.317045Z","signature_b64":"HtBSl6vuZNSyaVD6yX3i/BPv4mVz7jXn4rkPteM7Yg4G0tmDMmhM4wRn6A4wil1vrHuN0CZgw/86psLz/4YQCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e4dbb59b030c90a3476703e51bd1d1c116524ebdd154479f31ab1c90fe829049","last_reissued_at":"2026-07-05T06:09:09.316515Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:09:09.316515Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Solving Regularized Exp, Cosh and Sinh Regression Problems","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Tianyi Zhou, Zhao Song, Zhihang Li","submitted_at":"2023-03-28T04:26:51Z","abstract_excerpt":"In modern machine learning, attention computation is a fundamental task for training large language models such as Transformer, GPT-4 and ChatGPT. In this work, we study exponential regression problem which is inspired by the softmax/exp unit in the attention mechanism in large language models. The standard exponential regression is non-convex. We study the regularization version of exponential regression problem which is a convex problem. We use approximate newton method to solve in input sparsity time.\n  Formally, in this problem, one is given matrix $A \\in \\mathbb{R}^{n \\times d}$, $b \\in \\"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.15725","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2303.15725/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2303.15725","created_at":"2026-07-05T06:09:09.316585+00:00"},{"alias_kind":"arxiv_version","alias_value":"2303.15725v2","created_at":"2026-07-05T06:09:09.316585+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.15725","created_at":"2026-07-05T06:09:09.316585+00:00"},{"alias_kind":"pith_short_12","alias_value":"4TN3LGYDBSIK","created_at":"2026-07-05T06:09:09.316585+00:00"},{"alias_kind":"pith_short_16","alias_value":"4TN3LGYDBSIKGR3H","created_at":"2026-07-05T06:09:09.316585+00:00"},{"alias_kind":"pith_short_8","alias_value":"4TN3LGYD","created_at":"2026-07-05T06:09:09.316585+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2306.14048","citing_title":"H$_2$O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Models","ref_index":104,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4TN3LGYDBSIKGR3HAPSRXUORYE","json":"https://pith.science/pith/4TN3LGYDBSIKGR3HAPSRXUORYE.json","graph_json":"https://pith.science/api/pith-number/4TN3LGYDBSIKGR3HAPSRXUORYE/graph.json","events_json":"https://pith.science/api/pith-number/4TN3LGYDBSIKGR3HAPSRXUORYE/events.json","paper":"https://pith.science/paper/4TN3LGYD"},"agent_actions":{"view_html":"https://pith.science/pith/4TN3LGYDBSIKGR3HAPSRXUORYE","download_json":"https://pith.science/pith/4TN3LGYDBSIKGR3HAPSRXUORYE.json","view_paper":"https://pith.science/paper/4TN3LGYD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2303.15725&json=true","fetch_graph":"https://pith.science/api/pith-number/4TN3LGYDBSIKGR3HAPSRXUORYE/graph.json","fetch_events":"https://pith.science/api/pith-number/4TN3LGYDBSIKGR3HAPSRXUORYE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4TN3LGYDBSIKGR3HAPSRXUORYE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4TN3LGYDBSIKGR3HAPSRXUORYE/action/storage_attestation","attest_author":"https://pith.science/pith/4TN3LGYDBSIKGR3HAPSRXUORYE/action/author_attestation","sign_citation":"https://pith.science/pith/4TN3LGYDBSIKGR3HAPSRXUORYE/action/citation_signature","submit_replication":"https://pith.science/pith/4TN3LGYDBSIKGR3HAPSRXUORYE/action/replication_record"}},"created_at":"2026-07-05T06:09:09.316585+00:00","updated_at":"2026-07-05T06:09:09.316585+00:00"}