{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:54X65TVMHEYZN5H7O7MRKGSDBD","short_pith_number":"pith:54X65TVM","canonical_record":{"source":{"id":"2312.09979","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-12-15T17:45:06Z","cross_cats_sorted":[],"title_canon_sha256":"35cb8c3a5e678b007a08c91d798cae2c7b9ea459b4850da89ba245b857618767","abstract_canon_sha256":"f4e9becaf95f09211f145646d9326cfcce60ca4728e6b7f5ffb17b9f53cf65eb"},"schema_version":"1.0"},"canonical_sha256":"ef2feeceac393196f4ff77d9151a4308cdb3d1efe71dc872b76f7dfcda21d4ce","source":{"kind":"arxiv","id":"2312.09979","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2312.09979","created_at":"2026-07-05T07:53:49Z"},{"alias_kind":"arxiv_version","alias_value":"2312.09979v4","created_at":"2026-07-05T07:53:49Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.09979","created_at":"2026-07-05T07:53:49Z"},{"alias_kind":"pith_short_12","alias_value":"54X65TVMHEYZ","created_at":"2026-07-05T07:53:49Z"},{"alias_kind":"pith_short_16","alias_value":"54X65TVMHEYZN5H7","created_at":"2026-07-05T07:53:49Z"},{"alias_kind":"pith_short_8","alias_value":"54X65TVM","created_at":"2026-07-05T07:53:49Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:54X65TVMHEYZN5H7O7MRKGSDBD","target":"record","payload":{"canonical_record":{"source":{"id":"2312.09979","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-12-15T17:45:06Z","cross_cats_sorted":[],"title_canon_sha256":"35cb8c3a5e678b007a08c91d798cae2c7b9ea459b4850da89ba245b857618767","abstract_canon_sha256":"f4e9becaf95f09211f145646d9326cfcce60ca4728e6b7f5ffb17b9f53cf65eb"},"schema_version":"1.0"},"canonical_sha256":"ef2feeceac393196f4ff77d9151a4308cdb3d1efe71dc872b76f7dfcda21d4ce","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:53:49.786878Z","signature_b64":"oT32TTifPp4BxWWgualbChP42Vu8egpGbnpnn+IHVsfa9pfoy1Q6TVd7v1LOpxxEQ+hkvXgVNf11mIchE+XKCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ef2feeceac393196f4ff77d9151a4308cdb3d1efe71dc872b76f7dfcda21d4ce","last_reissued_at":"2026-07-05T07:53:49.786412Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:53:49.786412Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2312.09979","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:53:49Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"i0yeyiKdILuEFUbeNobq4ftrNGIvEC/ETlURi0Bqiy836HXx/VQ4uxhnLxPDG+6l/IPfpv3Zk/pdu5V3wxJeBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T16:18:32.036652Z"},"content_sha256":"c32c198adc3ad82133850f69395b70af71ba97c780c4243d63f9b608719e56fd","schema_version":"1.0","event_id":"sha256:c32c198adc3ad82133850f69395b70af71ba97c780c4243d63f9b608719e56fd"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:54X65TVMHEYZN5H7O7MRKGSDBD","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"LoRAMoE: Alleviate World Knowledge Forgetting in Large Language Models via MoE-Style Plugin","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Enyu Zhou, Jiang Zhu, Jun Zhao, Qi Zhang, Rui Zheng, Shihan Dou, Shiliang Pu, Songyang Gao, Tao Gui, Wei Shen, Xiaoran Fan, Xiao Wang, Xuanjing Huang, Yan Liu, Yuhao Zhou, Zhiheng Xi","submitted_at":"2023-12-15T17:45:06Z","abstract_excerpt":"Supervised fine-tuning (SFT) is a crucial step for large language models (LLMs), enabling them to align with human instructions and enhance their capabilities in downstream tasks. Increasing instruction data substantially is a direct solution to align the model with a broader range of downstream tasks or notably improve its performance on a specific task. However, we find that large-scale increases in instruction data can damage the world knowledge previously stored in LLMs. To address this challenge, we propose LoRAMoE, a novelty framework that introduces several low-rank adapters (LoRA) and "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.09979","kind":"arxiv","version":4},"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/2312.09979/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":"2026-07-05T07:53:49Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zq4UKn6uCWIXDTGopO+7AH9acMM9T4xyZF21ijQlCjswsqKRVefqi+Por9g4zvRqFbwFpM0gJ4akWXTfdAFPDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T16:18:32.037400Z"},"content_sha256":"8f3535145854b6867777d8440cdfb8efa99c7a770241522b869dbb539bc8e519","schema_version":"1.0","event_id":"sha256:8f3535145854b6867777d8440cdfb8efa99c7a770241522b869dbb539bc8e519"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/54X65TVMHEYZN5H7O7MRKGSDBD/bundle.json","state_url":"https://pith.science/pith/54X65TVMHEYZN5H7O7MRKGSDBD/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/54X65TVMHEYZN5H7O7MRKGSDBD/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":"2026-08-20T16:18:32Z","links":{"resolver":"https://pith.science/pith/54X65TVMHEYZN5H7O7MRKGSDBD","bundle":"https://pith.science/pith/54X65TVMHEYZN5H7O7MRKGSDBD/bundle.json","state":"https://pith.science/pith/54X65TVMHEYZN5H7O7MRKGSDBD/state.json","well_known_bundle":"https://pith.science/.well-known/pith/54X65TVMHEYZN5H7O7MRKGSDBD/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:54X65TVMHEYZN5H7O7MRKGSDBD","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":"f4e9becaf95f09211f145646d9326cfcce60ca4728e6b7f5ffb17b9f53cf65eb","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-12-15T17:45:06Z","title_canon_sha256":"35cb8c3a5e678b007a08c91d798cae2c7b9ea459b4850da89ba245b857618767"},"schema_version":"1.0","source":{"id":"2312.09979","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2312.09979","created_at":"2026-07-05T07:53:49Z"},{"alias_kind":"arxiv_version","alias_value":"2312.09979v4","created_at":"2026-07-05T07:53:49Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.09979","created_at":"2026-07-05T07:53:49Z"},{"alias_kind":"pith_short_12","alias_value":"54X65TVMHEYZ","created_at":"2026-07-05T07:53:49Z"},{"alias_kind":"pith_short_16","alias_value":"54X65TVMHEYZN5H7","created_at":"2026-07-05T07:53:49Z"},{"alias_kind":"pith_short_8","alias_value":"54X65TVM","created_at":"2026-07-05T07:53:49Z"}],"graph_snapshots":[{"event_id":"sha256:8f3535145854b6867777d8440cdfb8efa99c7a770241522b869dbb539bc8e519","target":"graph","created_at":"2026-07-05T07:53:49Z","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/2312.09979/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Supervised fine-tuning (SFT) is a crucial step for large language models (LLMs), enabling them to align with human instructions and enhance their capabilities in downstream tasks. Increasing instruction data substantially is a direct solution to align the model with a broader range of downstream tasks or notably improve its performance on a specific task. However, we find that large-scale increases in instruction data can damage the world knowledge previously stored in LLMs. To address this challenge, we propose LoRAMoE, a novelty framework that introduces several low-rank adapters (LoRA) and ","authors_text":"Enyu Zhou, Jiang Zhu, Jun Zhao, Qi Zhang, Rui Zheng, Shihan Dou, Shiliang Pu, Songyang Gao, Tao Gui, Wei Shen, Xiaoran Fan, Xiao Wang, Xuanjing Huang, Yan Liu, Yuhao Zhou, Zhiheng Xi","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-12-15T17:45:06Z","title":"LoRAMoE: Alleviate World Knowledge Forgetting in Large Language Models via MoE-Style Plugin"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.09979","kind":"arxiv","version":4},"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:c32c198adc3ad82133850f69395b70af71ba97c780c4243d63f9b608719e56fd","target":"record","created_at":"2026-07-05T07:53:49Z","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":"f4e9becaf95f09211f145646d9326cfcce60ca4728e6b7f5ffb17b9f53cf65eb","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-12-15T17:45:06Z","title_canon_sha256":"35cb8c3a5e678b007a08c91d798cae2c7b9ea459b4850da89ba245b857618767"},"schema_version":"1.0","source":{"id":"2312.09979","kind":"arxiv","version":4}},"canonical_sha256":"ef2feeceac393196f4ff77d9151a4308cdb3d1efe71dc872b76f7dfcda21d4ce","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ef2feeceac393196f4ff77d9151a4308cdb3d1efe71dc872b76f7dfcda21d4ce","first_computed_at":"2026-07-05T07:53:49.786412Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:53:49.786412Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"oT32TTifPp4BxWWgualbChP42Vu8egpGbnpnn+IHVsfa9pfoy1Q6TVd7v1LOpxxEQ+hkvXgVNf11mIchE+XKCw==","signature_status":"signed_v1","signed_at":"2026-07-05T07:53:49.786878Z","signed_message":"canonical_sha256_bytes"},"source_id":"2312.09979","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c32c198adc3ad82133850f69395b70af71ba97c780c4243d63f9b608719e56fd","sha256:8f3535145854b6867777d8440cdfb8efa99c7a770241522b869dbb539bc8e519"],"state_sha256":"564c8a751c7a01fb6aee37ab924c34e8fbc740395e41230a27da80562abba6a2"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"I+ezIWhAT8TlJla7yl33wbHSFs4kvIcMoCGN9/rOBuhPKhT111PqoQxinSYciReLdU5HQI0MfyXJl8qXo9ZGDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-20T16:18:32.046683Z","bundle_sha256":"a356c0ed9891bac90d0329f7c8a12063f03f4b518e460b77de99b6f12f85c0fd"}}