{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:DRI6FD436WI43OPHE5QRXTKKT3","short_pith_number":"pith:DRI6FD43","canonical_record":{"source":{"id":"2403.01244","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-02T16:11:23Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"b700330c759a53c57f49c4efdfe907139865953ceb0897ce36a7bb2ec418da11","abstract_canon_sha256":"518e3c84e4ec7258fe9782f70b514d3bc7db83111c39aefe6ae45b08bfbc51f9"},"schema_version":"1.0"},"canonical_sha256":"1c51e28f9bf591cdb9e727611bcd4a9ed6216868fdb1a3053de2fada4cf203a4","source":{"kind":"arxiv","id":"2403.01244","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.01244","created_at":"2026-07-05T08:23:19Z"},{"alias_kind":"arxiv_version","alias_value":"2403.01244v2","created_at":"2026-07-05T08:23:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.01244","created_at":"2026-07-05T08:23:19Z"},{"alias_kind":"pith_short_12","alias_value":"DRI6FD436WI4","created_at":"2026-07-05T08:23:19Z"},{"alias_kind":"pith_short_16","alias_value":"DRI6FD436WI43OPH","created_at":"2026-07-05T08:23:19Z"},{"alias_kind":"pith_short_8","alias_value":"DRI6FD43","created_at":"2026-07-05T08:23:19Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:DRI6FD436WI43OPHE5QRXTKKT3","target":"record","payload":{"canonical_record":{"source":{"id":"2403.01244","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-02T16:11:23Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"b700330c759a53c57f49c4efdfe907139865953ceb0897ce36a7bb2ec418da11","abstract_canon_sha256":"518e3c84e4ec7258fe9782f70b514d3bc7db83111c39aefe6ae45b08bfbc51f9"},"schema_version":"1.0"},"canonical_sha256":"1c51e28f9bf591cdb9e727611bcd4a9ed6216868fdb1a3053de2fada4cf203a4","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:23:19.338542Z","signature_b64":"WJDMhR/bJaStpfr/f0MgXzwjGR5L/uWjVFmdSzPmXgKHMRM3OFqr6czyln1L8gXJs9bynhZh8PDbCWI8/1nAAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1c51e28f9bf591cdb9e727611bcd4a9ed6216868fdb1a3053de2fada4cf203a4","last_reissued_at":"2026-07-05T08:23:19.338039Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:23:19.338039Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2403.01244","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":"2026-07-05T08:23:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"wqZEIzEcoI13UuEQpfcGQ02wUP8wZeSJTRifcw9MFxlXPPez1mYch2+RkBgkURzIX7cUV7y16sVfg86VETfdBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T14:05:35.367270Z"},"content_sha256":"ab63d0f7d6ea40f8eb66e329e2ff64be1096f2a3c1049df88c0274e3994d58e9","schema_version":"1.0","event_id":"sha256:ab63d0f7d6ea40f8eb66e329e2ff64be1096f2a3c1049df88c0274e3994d58e9"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:DRI6FD436WI43OPHE5QRXTKKT3","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Mitigating Catastrophic Forgetting in Large Language Models with Self-Synthesized Rehearsal","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ante Wang, Chengyi Yang, Jianheng Huang, Jinsong Su, Junfeng Yao, Leyang Cui, Linfeng Song, Xinting Liao","submitted_at":"2024-03-02T16:11:23Z","abstract_excerpt":"Large language models (LLMs) suffer from catastrophic forgetting during continual learning. Conventional rehearsal-based methods rely on previous training data to retain the model's ability, which may not be feasible in real-world applications. When conducting continual learning based on a publicly-released LLM checkpoint, the availability of the original training data may be non-existent. To address this challenge, we propose a framework called Self-Synthesized Rehearsal (SSR) that uses the LLM to generate synthetic instances for rehearsal. Concretely, we first employ the base LLM for in-cont"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.01244","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/2403.01244/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":"2026-07-05T08:23:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kQcsiQ9/eVpqk0hDuIiz2AsTda+JM5ewNj9F4v4JiKpbac9wd+71ul8PMdu/N1MmQM8LWAMhfpUiKqFtbem6Bw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T14:05:35.369416Z"},"content_sha256":"cf9abab9494f3dbb3f01c9bfdaaa79d8a27b90e74d575b9fd8d390f5fc57938a","schema_version":"1.0","event_id":"sha256:cf9abab9494f3dbb3f01c9bfdaaa79d8a27b90e74d575b9fd8d390f5fc57938a"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DRI6FD436WI43OPHE5QRXTKKT3/bundle.json","state_url":"https://pith.science/pith/DRI6FD436WI43OPHE5QRXTKKT3/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DRI6FD436WI43OPHE5QRXTKKT3/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":"2026-08-07T14:05:35Z","links":{"resolver":"https://pith.science/pith/DRI6FD436WI43OPHE5QRXTKKT3","bundle":"https://pith.science/pith/DRI6FD436WI43OPHE5QRXTKKT3/bundle.json","state":"https://pith.science/pith/DRI6FD436WI43OPHE5QRXTKKT3/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DRI6FD436WI43OPHE5QRXTKKT3/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:DRI6FD436WI43OPHE5QRXTKKT3","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":"518e3c84e4ec7258fe9782f70b514d3bc7db83111c39aefe6ae45b08bfbc51f9","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-02T16:11:23Z","title_canon_sha256":"b700330c759a53c57f49c4efdfe907139865953ceb0897ce36a7bb2ec418da11"},"schema_version":"1.0","source":{"id":"2403.01244","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.01244","created_at":"2026-07-05T08:23:19Z"},{"alias_kind":"arxiv_version","alias_value":"2403.01244v2","created_at":"2026-07-05T08:23:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.01244","created_at":"2026-07-05T08:23:19Z"},{"alias_kind":"pith_short_12","alias_value":"DRI6FD436WI4","created_at":"2026-07-05T08:23:19Z"},{"alias_kind":"pith_short_16","alias_value":"DRI6FD436WI43OPH","created_at":"2026-07-05T08:23:19Z"},{"alias_kind":"pith_short_8","alias_value":"DRI6FD43","created_at":"2026-07-05T08:23:19Z"}],"graph_snapshots":[{"event_id":"sha256:cf9abab9494f3dbb3f01c9bfdaaa79d8a27b90e74d575b9fd8d390f5fc57938a","target":"graph","created_at":"2026-07-05T08:23:19Z","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/2403.01244/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) suffer from catastrophic forgetting during continual learning. Conventional rehearsal-based methods rely on previous training data to retain the model's ability, which may not be feasible in real-world applications. When conducting continual learning based on a publicly-released LLM checkpoint, the availability of the original training data may be non-existent. To address this challenge, we propose a framework called Self-Synthesized Rehearsal (SSR) that uses the LLM to generate synthetic instances for rehearsal. Concretely, we first employ the base LLM for in-cont","authors_text":"Ante Wang, Chengyi Yang, Jianheng Huang, Jinsong Su, Junfeng Yao, Leyang Cui, Linfeng Song, Xinting Liao","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-02T16:11:23Z","title":"Mitigating Catastrophic Forgetting in Large Language Models with Self-Synthesized Rehearsal"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.01244","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:ab63d0f7d6ea40f8eb66e329e2ff64be1096f2a3c1049df88c0274e3994d58e9","target":"record","created_at":"2026-07-05T08:23:19Z","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":"518e3c84e4ec7258fe9782f70b514d3bc7db83111c39aefe6ae45b08bfbc51f9","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-02T16:11:23Z","title_canon_sha256":"b700330c759a53c57f49c4efdfe907139865953ceb0897ce36a7bb2ec418da11"},"schema_version":"1.0","source":{"id":"2403.01244","kind":"arxiv","version":2}},"canonical_sha256":"1c51e28f9bf591cdb9e727611bcd4a9ed6216868fdb1a3053de2fada4cf203a4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"1c51e28f9bf591cdb9e727611bcd4a9ed6216868fdb1a3053de2fada4cf203a4","first_computed_at":"2026-07-05T08:23:19.338039Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:23:19.338039Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"WJDMhR/bJaStpfr/f0MgXzwjGR5L/uWjVFmdSzPmXgKHMRM3OFqr6czyln1L8gXJs9bynhZh8PDbCWI8/1nAAA==","signature_status":"signed_v1","signed_at":"2026-07-05T08:23:19.338542Z","signed_message":"canonical_sha256_bytes"},"source_id":"2403.01244","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ab63d0f7d6ea40f8eb66e329e2ff64be1096f2a3c1049df88c0274e3994d58e9","sha256:cf9abab9494f3dbb3f01c9bfdaaa79d8a27b90e74d575b9fd8d390f5fc57938a"],"state_sha256":"15736384d680c96a644066813da83c768ec9f1732d8f30ef422618812c42191b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"urxzd2lMb7UARXqbO+cqxpuTdwFs5aud1S9HzN/SeK8E20NIlGvDlTFazZPI5TFA6JKC3pbJvM4sqrQOZvTDCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T14:05:35.381670Z","bundle_sha256":"ab6a2ddf1ee3b7ff536b04bda88e31ea100a14ef982a21054547ced5c4539684"}}