{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:O2N5G6VU2IAO32PPYJ4KQARJTQ","short_pith_number":"pith:O2N5G6VU","schema_version":"1.0","canonical_sha256":"769bd37ab4d200ede9efc278a802299c1d54d405e56335508faacca39ef3186a","source":{"kind":"arxiv","id":"2605.27591","version":1},"attestation_state":"computed","paper":{"title":"Gradient Transformer: Learning to Generate Updates for LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Binh-Nguyen Nguyen, Issa Khalil, Khang Tran, NHatHai Phan","submitted_at":"2026-05-26T19:05:13Z","abstract_excerpt":"Many organizations lack computational resources to fine-tune large language models (LLMs) on private (unshareable) data for better utility, while fine-tuning tiny language models (TinyLMs) alone performs poorly. To address this bottleneck, we propose a data-free knowledge distillation framework that generates LLM update vectors based on TinyLMs fine-tuned on private data. An update vector is a vector of parameter changes from an initial model to its fine-tuned version on a dataset, capturing the effect of cumulative gradient steps during fine-tuning. The key idea of our framework is a novel Gr"},"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":"2605.27591","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-26T19:05:13Z","cross_cats_sorted":[],"title_canon_sha256":"78c0f65ad712236d7452aaabb96d4ed445b9cfcfa5ac368340eac6f82cb6d93d","abstract_canon_sha256":"8a6381071585008f6da6f218d2dd975a151d12a5a7d71640bb58ad11143301f7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-28T01:04:17.159506Z","signature_b64":"4AIGYOX2P90/XWA9DdhhRRfv6/9EM+tPygMUFRKJKYHJFXoOKwlOV12LT9n9DARx5txihPVTzuD8JyAHxLM6Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"769bd37ab4d200ede9efc278a802299c1d54d405e56335508faacca39ef3186a","last_reissued_at":"2026-05-28T01:04:17.159033Z","signature_status":"signed_v1","first_computed_at":"2026-05-28T01:04:17.159033Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Gradient Transformer: Learning to Generate Updates for LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Binh-Nguyen Nguyen, Issa Khalil, Khang Tran, NHatHai Phan","submitted_at":"2026-05-26T19:05:13Z","abstract_excerpt":"Many organizations lack computational resources to fine-tune large language models (LLMs) on private (unshareable) data for better utility, while fine-tuning tiny language models (TinyLMs) alone performs poorly. To address this bottleneck, we propose a data-free knowledge distillation framework that generates LLM update vectors based on TinyLMs fine-tuned on private data. An update vector is a vector of parameter changes from an initial model to its fine-tuned version on a dataset, capturing the effect of cumulative gradient steps during fine-tuning. The key idea of our framework is a novel Gr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.27591","kind":"arxiv","version":1},"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/2605.27591/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":"2605.27591","created_at":"2026-05-28T01:04:17.159087+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.27591v1","created_at":"2026-05-28T01:04:17.159087+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.27591","created_at":"2026-05-28T01:04:17.159087+00:00"},{"alias_kind":"pith_short_12","alias_value":"O2N5G6VU2IAO","created_at":"2026-05-28T01:04:17.159087+00:00"},{"alias_kind":"pith_short_16","alias_value":"O2N5G6VU2IAO32PP","created_at":"2026-05-28T01:04:17.159087+00:00"},{"alias_kind":"pith_short_8","alias_value":"O2N5G6VU","created_at":"2026-05-28T01:04:17.159087+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/O2N5G6VU2IAO32PPYJ4KQARJTQ","json":"https://pith.science/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ.json","graph_json":"https://pith.science/api/pith-number/O2N5G6VU2IAO32PPYJ4KQARJTQ/graph.json","events_json":"https://pith.science/api/pith-number/O2N5G6VU2IAO32PPYJ4KQARJTQ/events.json","paper":"https://pith.science/paper/O2N5G6VU"},"agent_actions":{"view_html":"https://pith.science/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ","download_json":"https://pith.science/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ.json","view_paper":"https://pith.science/paper/O2N5G6VU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.27591&json=true","fetch_graph":"https://pith.science/api/pith-number/O2N5G6VU2IAO32PPYJ4KQARJTQ/graph.json","fetch_events":"https://pith.science/api/pith-number/O2N5G6VU2IAO32PPYJ4KQARJTQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ/action/storage_attestation","attest_author":"https://pith.science/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ/action/author_attestation","sign_citation":"https://pith.science/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ/action/citation_signature","submit_replication":"https://pith.science/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ/action/replication_record"}},"created_at":"2026-05-28T01:04:17.159087+00:00","updated_at":"2026-05-28T01:04:17.159087+00:00"}