{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:O2N5G6VU2IAO32PPYJ4KQARJTQ","short_pith_number":"pith:O2N5G6VU","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"},"canonical_sha256":"769bd37ab4d200ede9efc278a802299c1d54d405e56335508faacca39ef3186a","source":{"kind":"arxiv","id":"2605.27591","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.27591","created_at":"2026-05-28T01:04:17Z"},{"alias_kind":"arxiv_version","alias_value":"2605.27591v1","created_at":"2026-05-28T01:04:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.27591","created_at":"2026-05-28T01:04:17Z"},{"alias_kind":"pith_short_12","alias_value":"O2N5G6VU2IAO","created_at":"2026-05-28T01:04:17Z"},{"alias_kind":"pith_short_16","alias_value":"O2N5G6VU2IAO32PP","created_at":"2026-05-28T01:04:17Z"},{"alias_kind":"pith_short_8","alias_value":"O2N5G6VU","created_at":"2026-05-28T01:04:17Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:O2N5G6VU2IAO32PPYJ4KQARJTQ","target":"record","payload":{"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"},"canonical_sha256":"769bd37ab4d200ede9efc278a802299c1d54d405e56335508faacca39ef3186a","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"},"source_kind":"arxiv","source_id":"2605.27591","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-28T01:04:17Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"cZny9Z9JW/SHlQsQ0IebDARLLxSiM7A09OuWLOq1ccVefnAW62BjtMd3oa08TcEbIYSgOQH8NTkopS558D30Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-01T21:12:32.741015Z"},"content_sha256":"f5d0a7c219ad20af2a4e10744fac325a2298e673190f798dff7ca4bb99d049f9","schema_version":"1.0","event_id":"sha256:f5d0a7c219ad20af2a4e10744fac325a2298e673190f798dff7ca4bb99d049f9"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:O2N5G6VU2IAO32PPYJ4KQARJTQ","target":"graph","payload":{"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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-28T01:04:17Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"pgBBeIeCPKMvJoCYjCRSjTP21vlMLF2fHeLk4DU7GPZ/Y8T2mw2qI0+wYyYhL3WH5vfUdGKiK8XFYlmpCxZjDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-01T21:12:32.741398Z"},"content_sha256":"bbfa93d47796ed7a103f91865506f9c7784883dacd297828e26399b046d15a74","schema_version":"1.0","event_id":"sha256:bbfa93d47796ed7a103f91865506f9c7784883dacd297828e26399b046d15a74"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ/bundle.json","state_url":"https://pith.science/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ/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":"2026-06-01T21:12:32Z","links":{"resolver":"https://pith.science/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ","bundle":"https://pith.science/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ/bundle.json","state":"https://pith.science/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/O2N5G6VU2IAO32PPYJ4KQARJTQ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:O2N5G6VU2IAO32PPYJ4KQARJTQ","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":"8a6381071585008f6da6f218d2dd975a151d12a5a7d71640bb58ad11143301f7","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-26T19:05:13Z","title_canon_sha256":"78c0f65ad712236d7452aaabb96d4ed445b9cfcfa5ac368340eac6f82cb6d93d"},"schema_version":"1.0","source":{"id":"2605.27591","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.27591","created_at":"2026-05-28T01:04:17Z"},{"alias_kind":"arxiv_version","alias_value":"2605.27591v1","created_at":"2026-05-28T01:04:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.27591","created_at":"2026-05-28T01:04:17Z"},{"alias_kind":"pith_short_12","alias_value":"O2N5G6VU2IAO","created_at":"2026-05-28T01:04:17Z"},{"alias_kind":"pith_short_16","alias_value":"O2N5G6VU2IAO32PP","created_at":"2026-05-28T01:04:17Z"},{"alias_kind":"pith_short_8","alias_value":"O2N5G6VU","created_at":"2026-05-28T01:04:17Z"}],"graph_snapshots":[{"event_id":"sha256:bbfa93d47796ed7a103f91865506f9c7784883dacd297828e26399b046d15a74","target":"graph","created_at":"2026-05-28T01:04:17Z","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/2605.27591/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"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","authors_text":"Binh-Nguyen Nguyen, Issa Khalil, Khang Tran, NHatHai Phan","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-26T19:05:13Z","title":"Gradient Transformer: Learning to Generate Updates for LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.27591","kind":"arxiv","version":1},"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:f5d0a7c219ad20af2a4e10744fac325a2298e673190f798dff7ca4bb99d049f9","target":"record","created_at":"2026-05-28T01:04:17Z","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":"8a6381071585008f6da6f218d2dd975a151d12a5a7d71640bb58ad11143301f7","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-26T19:05:13Z","title_canon_sha256":"78c0f65ad712236d7452aaabb96d4ed445b9cfcfa5ac368340eac6f82cb6d93d"},"schema_version":"1.0","source":{"id":"2605.27591","kind":"arxiv","version":1}},"canonical_sha256":"769bd37ab4d200ede9efc278a802299c1d54d405e56335508faacca39ef3186a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"769bd37ab4d200ede9efc278a802299c1d54d405e56335508faacca39ef3186a","first_computed_at":"2026-05-28T01:04:17.159033Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-28T01:04:17.159033Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4AIGYOX2P90/XWA9DdhhRRfv6/9EM+tPygMUFRKJKYHJFXoOKwlOV12LT9n9DARx5txihPVTzuD8JyAHxLM6Dg==","signature_status":"signed_v1","signed_at":"2026-05-28T01:04:17.159506Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.27591","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f5d0a7c219ad20af2a4e10744fac325a2298e673190f798dff7ca4bb99d049f9","sha256:bbfa93d47796ed7a103f91865506f9c7784883dacd297828e26399b046d15a74"],"state_sha256":"f7c47758ddff10a8e2a8b28e3cf77f6ec4789fc9f9abe53154befb58fdf5b829"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PFpwQQm30ozxvEz33R/6R67IvhQUO5v9CjOOVR9/lavVjZba9UdRGX5+HX9tO7qdVDWe/CVno13RiLbUoz3+Dw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-01T21:12:32.744238Z","bundle_sha256":"962cfacb306f38a94e2d949b1c31464b853560fe2337fe13291d066e2c580b55"}}