{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:ODQUDRQ5TYNFRSYCPUPHLMYD45","short_pith_number":"pith:ODQUDRQ5","canonical_record":{"source":{"id":"2502.13533","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-19T08:39:15Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"8b916e4a7a803d46f705de46f51fe3f4cd48a426208e93ceb9d25239789986b7","abstract_canon_sha256":"55386756f9f10e69e5314c13ac2db4aa8071e502781c4df1c16c202ca561cb4c"},"schema_version":"1.0"},"canonical_sha256":"70e141c61d9e1a58cb027d1e75b303e77d68ca3e8150a3fb626f7bfba3eac7a8","source":{"kind":"arxiv","id":"2502.13533","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.13533","created_at":"2026-07-05T10:31:54Z"},{"alias_kind":"arxiv_version","alias_value":"2502.13533v2","created_at":"2026-07-05T10:31:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.13533","created_at":"2026-07-05T10:31:54Z"},{"alias_kind":"pith_short_12","alias_value":"ODQUDRQ5TYNF","created_at":"2026-07-05T10:31:54Z"},{"alias_kind":"pith_short_16","alias_value":"ODQUDRQ5TYNFRSYC","created_at":"2026-07-05T10:31:54Z"},{"alias_kind":"pith_short_8","alias_value":"ODQUDRQ5","created_at":"2026-07-05T10:31:54Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:ODQUDRQ5TYNFRSYCPUPHLMYD45","target":"record","payload":{"canonical_record":{"source":{"id":"2502.13533","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-19T08:39:15Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"8b916e4a7a803d46f705de46f51fe3f4cd48a426208e93ceb9d25239789986b7","abstract_canon_sha256":"55386756f9f10e69e5314c13ac2db4aa8071e502781c4df1c16c202ca561cb4c"},"schema_version":"1.0"},"canonical_sha256":"70e141c61d9e1a58cb027d1e75b303e77d68ca3e8150a3fb626f7bfba3eac7a8","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:31:54.378414Z","signature_b64":"mbGiu2UCaz3qeQjJerAy+hKKtqQMrmfjvpJA3raeUbtuGsugPZm8wqM1vKMBiI1+KDBX7jYKd2raVIJzBl2wDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"70e141c61d9e1a58cb027d1e75b303e77d68ca3e8150a3fb626f7bfba3eac7a8","last_reissued_at":"2026-07-05T10:31:54.377787Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:31:54.377787Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.13533","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:31:54Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IqB23i9aiUhs+RKsdwGOByEjvWkR7djAbXYM+mJn6UdNF3arFFmHwoFk7QbQ+rFxMMYstsYg9Azbk9vMAnMaBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T08:24:35.704309Z"},"content_sha256":"51715a5d77120bd7e76245d7db14f35bebbc02710e9c70f09a62001cc9ae69b5","schema_version":"1.0","event_id":"sha256:51715a5d77120bd7e76245d7db14f35bebbc02710e9c70f09a62001cc9ae69b5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:ODQUDRQ5TYNFRSYCPUPHLMYD45","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Train Small, Infer Large: Memory-Efficient LoRA Training for Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Guiming Xie, Huan Li, Jue Wang, Jun Zhang, Ke Chen, Kunlong Zhou, Lidan Shou, Xuejian Gong, Yang You","submitted_at":"2025-02-19T08:39:15Z","abstract_excerpt":"Large Language Models (LLMs) have significantly advanced natural language processing with exceptional task generalization capabilities. Low-Rank Adaption (LoRA) offers a cost-effective fine-tuning solution, freezing the original model parameters and training only lightweight, low-rank adapter matrices. However, the memory footprint of LoRA is largely dominated by the original model parameters. To mitigate this, we propose LoRAM, a memory-efficient LoRA training scheme founded on the intuition that many neurons in over-parameterized LLMs have low training utility but are essential for inference"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.13533","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/2502.13533/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":"2026-07-05T10:31:54Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9xyew148uxFQ4gBpAzMZUAqBTNJjguJPRvTP5ljOwPwpj7v/8NW2AGfvn+xXV91PN7D5NMdZBfwjOTQJrrZ0Bg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T08:24:35.704707Z"},"content_sha256":"e9c477da2e1e2ee911659729d8a4cef695b48453a0198d80eabf9ba82be2f57b","schema_version":"1.0","event_id":"sha256:e9c477da2e1e2ee911659729d8a4cef695b48453a0198d80eabf9ba82be2f57b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ODQUDRQ5TYNFRSYCPUPHLMYD45/bundle.json","state_url":"https://pith.science/pith/ODQUDRQ5TYNFRSYCPUPHLMYD45/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ODQUDRQ5TYNFRSYCPUPHLMYD45/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":"2026-07-07T08:24:35Z","links":{"resolver":"https://pith.science/pith/ODQUDRQ5TYNFRSYCPUPHLMYD45","bundle":"https://pith.science/pith/ODQUDRQ5TYNFRSYCPUPHLMYD45/bundle.json","state":"https://pith.science/pith/ODQUDRQ5TYNFRSYCPUPHLMYD45/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ODQUDRQ5TYNFRSYCPUPHLMYD45/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:ODQUDRQ5TYNFRSYCPUPHLMYD45","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":"55386756f9f10e69e5314c13ac2db4aa8071e502781c4df1c16c202ca561cb4c","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-19T08:39:15Z","title_canon_sha256":"8b916e4a7a803d46f705de46f51fe3f4cd48a426208e93ceb9d25239789986b7"},"schema_version":"1.0","source":{"id":"2502.13533","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.13533","created_at":"2026-07-05T10:31:54Z"},{"alias_kind":"arxiv_version","alias_value":"2502.13533v2","created_at":"2026-07-05T10:31:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.13533","created_at":"2026-07-05T10:31:54Z"},{"alias_kind":"pith_short_12","alias_value":"ODQUDRQ5TYNF","created_at":"2026-07-05T10:31:54Z"},{"alias_kind":"pith_short_16","alias_value":"ODQUDRQ5TYNFRSYC","created_at":"2026-07-05T10:31:54Z"},{"alias_kind":"pith_short_8","alias_value":"ODQUDRQ5","created_at":"2026-07-05T10:31:54Z"}],"graph_snapshots":[{"event_id":"sha256:e9c477da2e1e2ee911659729d8a4cef695b48453a0198d80eabf9ba82be2f57b","target":"graph","created_at":"2026-07-05T10:31:54Z","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/2502.13533/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) have significantly advanced natural language processing with exceptional task generalization capabilities. Low-Rank Adaption (LoRA) offers a cost-effective fine-tuning solution, freezing the original model parameters and training only lightweight, low-rank adapter matrices. However, the memory footprint of LoRA is largely dominated by the original model parameters. To mitigate this, we propose LoRAM, a memory-efficient LoRA training scheme founded on the intuition that many neurons in over-parameterized LLMs have low training utility but are essential for inference","authors_text":"Guiming Xie, Huan Li, Jue Wang, Jun Zhang, Ke Chen, Kunlong Zhou, Lidan Shou, Xuejian Gong, Yang You","cross_cats":["cs.AI","cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-19T08:39:15Z","title":"Train Small, Infer Large: Memory-Efficient LoRA Training for Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.13533","kind":"arxiv","version":2},"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":"sha256:51715a5d77120bd7e76245d7db14f35bebbc02710e9c70f09a62001cc9ae69b5","target":"record","created_at":"2026-07-05T10:31:54Z","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":"55386756f9f10e69e5314c13ac2db4aa8071e502781c4df1c16c202ca561cb4c","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-19T08:39:15Z","title_canon_sha256":"8b916e4a7a803d46f705de46f51fe3f4cd48a426208e93ceb9d25239789986b7"},"schema_version":"1.0","source":{"id":"2502.13533","kind":"arxiv","version":2}},"canonical_sha256":"70e141c61d9e1a58cb027d1e75b303e77d68ca3e8150a3fb626f7bfba3eac7a8","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"70e141c61d9e1a58cb027d1e75b303e77d68ca3e8150a3fb626f7bfba3eac7a8","first_computed_at":"2026-07-05T10:31:54.377787Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:31:54.377787Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"mbGiu2UCaz3qeQjJerAy+hKKtqQMrmfjvpJA3raeUbtuGsugPZm8wqM1vKMBiI1+KDBX7jYKd2raVIJzBl2wDA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:31:54.378414Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.13533","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:51715a5d77120bd7e76245d7db14f35bebbc02710e9c70f09a62001cc9ae69b5","sha256:e9c477da2e1e2ee911659729d8a4cef695b48453a0198d80eabf9ba82be2f57b"],"state_sha256":"661de4eb6663e4ab0e8efa62e429e363d52f5a1c4a3ae4f6052afb48dcd3a296"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zGHXZaKexJemXMBmNqXcZ1dBDZZ7l7FRQXYL+90Sas/gxNF7UVqSwztRF0BukY5k1gJ6h257XP88KO24POj2Aw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T08:24:35.706697Z","bundle_sha256":"484652a3fda486d58cbf0b953a96e7bb06de7eed5a29eaadd480e7de2eef0da1"}}