{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:AUHZIGKLDFNMXI6YTBAWPG3UCB","short_pith_number":"pith:AUHZIGKL","canonical_record":{"source":{"id":"2501.13669","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-23T13:54:53Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"409f2eac9d5c8459428b6016a3282f1e6cd3dcbe74310aeaf4da5a576849b6c4","abstract_canon_sha256":"55d6642b9d96a36cbdd97e9373a9fa7003cb3cd17be35255285f3c8086286693"},"schema_version":"1.0"},"canonical_sha256":"050f94194b195acba3d89841679b74107842a1a040a9ce4c593390f313b32008","source":{"kind":"arxiv","id":"2501.13669","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.13669","created_at":"2026-07-05T10:15:24Z"},{"alias_kind":"arxiv_version","alias_value":"2501.13669v2","created_at":"2026-07-05T10:15:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.13669","created_at":"2026-07-05T10:15:24Z"},{"alias_kind":"pith_short_12","alias_value":"AUHZIGKLDFNM","created_at":"2026-07-05T10:15:24Z"},{"alias_kind":"pith_short_16","alias_value":"AUHZIGKLDFNMXI6Y","created_at":"2026-07-05T10:15:24Z"},{"alias_kind":"pith_short_8","alias_value":"AUHZIGKL","created_at":"2026-07-05T10:15:24Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:AUHZIGKLDFNMXI6YTBAWPG3UCB","target":"record","payload":{"canonical_record":{"source":{"id":"2501.13669","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-23T13:54:53Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"409f2eac9d5c8459428b6016a3282f1e6cd3dcbe74310aeaf4da5a576849b6c4","abstract_canon_sha256":"55d6642b9d96a36cbdd97e9373a9fa7003cb3cd17be35255285f3c8086286693"},"schema_version":"1.0"},"canonical_sha256":"050f94194b195acba3d89841679b74107842a1a040a9ce4c593390f313b32008","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:15:24.980936Z","signature_b64":"xO4Q+R65VxnEdSMq+ARoBAAzRL9uXgz9VVjoRELSpXy+9rmbsWdDHrtwscHUbcKjciN0vmqLrLdOcgYMYnFqBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"050f94194b195acba3d89841679b74107842a1a040a9ce4c593390f313b32008","last_reissued_at":"2026-07-05T10:15:24.980376Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:15:24.980376Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2501.13669","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":"2026-07-05T10:15:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"JDImf6rbZTdYq2HuJTkxq8mDxy4HCJeOFbi4Vb8RGxgm9ywCDzdk47/g1NBYmAY0n062WrbnZlqmMnjzVRQjDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T18:16:25.088437Z"},"content_sha256":"ec2a86d608cd8cdf3d004e7d61f5698979eed4710a6ca0d7bded442cf5d37e5b","schema_version":"1.0","event_id":"sha256:ec2a86d608cd8cdf3d004e7d61f5698979eed4710a6ca0d7bded442cf5d37e5b"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:AUHZIGKLDFNMXI6YTBAWPG3UCB","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"How to Alleviate Catastrophic Forgetting in LLMs Finetuning? Hierarchical Layer-Wise and Element-Wise Regularization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hao Xu, Jie Yu, Jun Ma, Long Peng, Qian Wan, Shasha Li, Shezheng Song, XiaoDong Liu","submitted_at":"2025-01-23T13:54:53Z","abstract_excerpt":"Large Language Models (LLMs) exhibit strong general language capabilities. However, fine-tuning these models on domain-specific tasks often leads to catastrophic forgetting, where the model overwrites or loses essential knowledge acquired during pretraining. This phenomenon significantly limits the broader applicability of LLMs. To address this challenge, we propose a novel approach to compute the element-wise importance of model parameters crucial for preserving general knowledge during fine-tuning. Our method utilizes a dual-objective optimization strategy: (1) regularization loss based on e"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.13669","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/2501.13669/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":"2026-07-05T10:15:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Uw2T6kpcv3HKiqJCw8os6Xt07biucBGyFsgYiEM2yqeCr/2g6JRAO8pSdOyceUCIKTfFHteQRZ8FdRZpD2/fDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T18:16:25.088934Z"},"content_sha256":"8663b80f3c71412b1295948539f5502813cfb52225d6883550799d54141ad4cf","schema_version":"1.0","event_id":"sha256:8663b80f3c71412b1295948539f5502813cfb52225d6883550799d54141ad4cf"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/AUHZIGKLDFNMXI6YTBAWPG3UCB/bundle.json","state_url":"https://pith.science/pith/AUHZIGKLDFNMXI6YTBAWPG3UCB/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/AUHZIGKLDFNMXI6YTBAWPG3UCB/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":"2026-08-17T18:16:25Z","links":{"resolver":"https://pith.science/pith/AUHZIGKLDFNMXI6YTBAWPG3UCB","bundle":"https://pith.science/pith/AUHZIGKLDFNMXI6YTBAWPG3UCB/bundle.json","state":"https://pith.science/pith/AUHZIGKLDFNMXI6YTBAWPG3UCB/state.json","well_known_bundle":"https://pith.science/.well-known/pith/AUHZIGKLDFNMXI6YTBAWPG3UCB/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:AUHZIGKLDFNMXI6YTBAWPG3UCB","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":"55d6642b9d96a36cbdd97e9373a9fa7003cb3cd17be35255285f3c8086286693","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-23T13:54:53Z","title_canon_sha256":"409f2eac9d5c8459428b6016a3282f1e6cd3dcbe74310aeaf4da5a576849b6c4"},"schema_version":"1.0","source":{"id":"2501.13669","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.13669","created_at":"2026-07-05T10:15:24Z"},{"alias_kind":"arxiv_version","alias_value":"2501.13669v2","created_at":"2026-07-05T10:15:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.13669","created_at":"2026-07-05T10:15:24Z"},{"alias_kind":"pith_short_12","alias_value":"AUHZIGKLDFNM","created_at":"2026-07-05T10:15:24Z"},{"alias_kind":"pith_short_16","alias_value":"AUHZIGKLDFNMXI6Y","created_at":"2026-07-05T10:15:24Z"},{"alias_kind":"pith_short_8","alias_value":"AUHZIGKL","created_at":"2026-07-05T10:15:24Z"}],"graph_snapshots":[{"event_id":"sha256:8663b80f3c71412b1295948539f5502813cfb52225d6883550799d54141ad4cf","target":"graph","created_at":"2026-07-05T10:15:24Z","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/2501.13669/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) exhibit strong general language capabilities. However, fine-tuning these models on domain-specific tasks often leads to catastrophic forgetting, where the model overwrites or loses essential knowledge acquired during pretraining. This phenomenon significantly limits the broader applicability of LLMs. To address this challenge, we propose a novel approach to compute the element-wise importance of model parameters crucial for preserving general knowledge during fine-tuning. Our method utilizes a dual-objective optimization strategy: (1) regularization loss based on e","authors_text":"Hao Xu, Jie Yu, Jun Ma, Long Peng, Qian Wan, Shasha Li, Shezheng Song, XiaoDong Liu","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-23T13:54:53Z","title":"How to Alleviate Catastrophic Forgetting in LLMs Finetuning? Hierarchical Layer-Wise and Element-Wise Regularization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.13669","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:ec2a86d608cd8cdf3d004e7d61f5698979eed4710a6ca0d7bded442cf5d37e5b","target":"record","created_at":"2026-07-05T10:15:24Z","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":"55d6642b9d96a36cbdd97e9373a9fa7003cb3cd17be35255285f3c8086286693","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-23T13:54:53Z","title_canon_sha256":"409f2eac9d5c8459428b6016a3282f1e6cd3dcbe74310aeaf4da5a576849b6c4"},"schema_version":"1.0","source":{"id":"2501.13669","kind":"arxiv","version":2}},"canonical_sha256":"050f94194b195acba3d89841679b74107842a1a040a9ce4c593390f313b32008","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"050f94194b195acba3d89841679b74107842a1a040a9ce4c593390f313b32008","first_computed_at":"2026-07-05T10:15:24.980376Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:15:24.980376Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"xO4Q+R65VxnEdSMq+ARoBAAzRL9uXgz9VVjoRELSpXy+9rmbsWdDHrtwscHUbcKjciN0vmqLrLdOcgYMYnFqBw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:15:24.980936Z","signed_message":"canonical_sha256_bytes"},"source_id":"2501.13669","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ec2a86d608cd8cdf3d004e7d61f5698979eed4710a6ca0d7bded442cf5d37e5b","sha256:8663b80f3c71412b1295948539f5502813cfb52225d6883550799d54141ad4cf"],"state_sha256":"b856e1af4f6999db1e5e3652837d6ae1f83e8cf5dff8d8ffcd1a7f0faa76e19c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tTO4JMoXPWKRklIFFjUFJBjKSoaXyGLFuvDqkAlr9iHx8aKfQ82NZkgtgmuSKoDwN20uQcMN11SKVIInpUI6DA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-17T18:16:25.092906Z","bundle_sha256":"7393d1003ab82624452b1bb39dc809585942336e2f46ec4ad342592cf6827a25"}}