{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:NADBFOOLYGJNGQAEI2IQGK5OVG","short_pith_number":"pith:NADBFOOL","canonical_record":{"source":{"id":"2505.12043","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-17T15:12:47Z","cross_cats_sorted":[],"title_canon_sha256":"2f24bfa8859bf9ca1ba1da9d02922bfe6fb99de5582c2c075d732c8a3a222e8e","abstract_canon_sha256":"9c13e3561321c32109d949b0b59f29adf95ef4e19e4d8ac0bb4d89d56379209a"},"schema_version":"1.0"},"canonical_sha256":"680612b9cbc192d340044691032baea991a71dd336d9ad2210dce4f4916f574c","source":{"kind":"arxiv","id":"2505.12043","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.12043","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"arxiv_version","alias_value":"2505.12043v2","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12043","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"pith_short_12","alias_value":"NADBFOOLYGJN","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"pith_short_16","alias_value":"NADBFOOLYGJNGQAE","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"pith_short_8","alias_value":"NADBFOOL","created_at":"2026-07-05T11:05:45Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:NADBFOOLYGJNGQAEI2IQGK5OVG","target":"record","payload":{"canonical_record":{"source":{"id":"2505.12043","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-17T15:12:47Z","cross_cats_sorted":[],"title_canon_sha256":"2f24bfa8859bf9ca1ba1da9d02922bfe6fb99de5582c2c075d732c8a3a222e8e","abstract_canon_sha256":"9c13e3561321c32109d949b0b59f29adf95ef4e19e4d8ac0bb4d89d56379209a"},"schema_version":"1.0"},"canonical_sha256":"680612b9cbc192d340044691032baea991a71dd336d9ad2210dce4f4916f574c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:05:45.160247Z","signature_b64":"qmXq6+EWur6F8Vnjz+Qqlldv43L1p+0/cCyxYDEEmgiRGJmPQcnCwtMfyn0v2B8IjjELlR7wUSzpet2Y3YdcCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"680612b9cbc192d340044691032baea991a71dd336d9ad2210dce4f4916f574c","last_reissued_at":"2026-07-05T11:05:45.159783Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:05:45.159783Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.12043","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":"2026-07-05T11:05:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KjKeng+vURlPaUJkR6q6gJl6Ul2keCVF5ZROX2gUsRrqXPc+SczMGSz0csmmWoZygi8O7MzVqFPMXfOJsKgeDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T22:54:40.500081Z"},"content_sha256":"5b1795d294f07587a5b81ae099587a8b14c893945da94afee2b75649e6fcd967","schema_version":"1.0","event_id":"sha256:5b1795d294f07587a5b81ae099587a8b14c893945da94afee2b75649e6fcd967"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:NADBFOOLYGJNGQAEI2IQGK5OVG","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"MoL for LLMs: Dual-Loss Optimization to Enhance Domain Expertise While Preserving General Capabilities","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jingxue Chen, Qianchun Lu, Qingkun Tang, Siyuan Fang","submitted_at":"2025-05-17T15:12:47Z","abstract_excerpt":"Although large language models (LLMs) perform well in general tasks, domain-specific applications suffer from hallucinations and accuracy limitations. Continual Pre-Training (CPT) approaches encounter two key issues: (1) domain-biased data degrades general language skills, and (2) improper corpus-mixture ratios limit effective adaptation. To address these, we propose a novel framework, Mixture of Losses (MoL), which decouples optimization objectives for domain-specific and general corpora. Specifically, cross-entropy (CE) loss is applied to domain-corpus to ensure knowledge acquisition, while "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12043","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/2505.12043/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":"2026-07-05T11:05:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2sZg6D5e4Ic9TX9vy0tfGnWszXCreOHz0YvTe/wQyQk4JNDEa6cwXndDp+/DKZSUr1v06s2D2S12rhAnmmsAAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T22:54:40.500860Z"},"content_sha256":"cd185b6884e4bf9b451d68ab09f62bbbf5d9e8a8087a894b1a337c48dc237710","schema_version":"1.0","event_id":"sha256:cd185b6884e4bf9b451d68ab09f62bbbf5d9e8a8087a894b1a337c48dc237710"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/NADBFOOLYGJNGQAEI2IQGK5OVG/bundle.json","state_url":"https://pith.science/pith/NADBFOOLYGJNGQAEI2IQGK5OVG/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/NADBFOOLYGJNGQAEI2IQGK5OVG/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":"2026-08-08T22:54:40Z","links":{"resolver":"https://pith.science/pith/NADBFOOLYGJNGQAEI2IQGK5OVG","bundle":"https://pith.science/pith/NADBFOOLYGJNGQAEI2IQGK5OVG/bundle.json","state":"https://pith.science/pith/NADBFOOLYGJNGQAEI2IQGK5OVG/state.json","well_known_bundle":"https://pith.science/.well-known/pith/NADBFOOLYGJNGQAEI2IQGK5OVG/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:NADBFOOLYGJNGQAEI2IQGK5OVG","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":"9c13e3561321c32109d949b0b59f29adf95ef4e19e4d8ac0bb4d89d56379209a","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-17T15:12:47Z","title_canon_sha256":"2f24bfa8859bf9ca1ba1da9d02922bfe6fb99de5582c2c075d732c8a3a222e8e"},"schema_version":"1.0","source":{"id":"2505.12043","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.12043","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"arxiv_version","alias_value":"2505.12043v2","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12043","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"pith_short_12","alias_value":"NADBFOOLYGJN","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"pith_short_16","alias_value":"NADBFOOLYGJNGQAE","created_at":"2026-07-05T11:05:45Z"},{"alias_kind":"pith_short_8","alias_value":"NADBFOOL","created_at":"2026-07-05T11:05:45Z"}],"graph_snapshots":[{"event_id":"sha256:cd185b6884e4bf9b451d68ab09f62bbbf5d9e8a8087a894b1a337c48dc237710","target":"graph","created_at":"2026-07-05T11:05:45Z","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/2505.12043/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Although large language models (LLMs) perform well in general tasks, domain-specific applications suffer from hallucinations and accuracy limitations. Continual Pre-Training (CPT) approaches encounter two key issues: (1) domain-biased data degrades general language skills, and (2) improper corpus-mixture ratios limit effective adaptation. To address these, we propose a novel framework, Mixture of Losses (MoL), which decouples optimization objectives for domain-specific and general corpora. Specifically, cross-entropy (CE) loss is applied to domain-corpus to ensure knowledge acquisition, while ","authors_text":"Jingxue Chen, Qianchun Lu, Qingkun Tang, Siyuan Fang","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-17T15:12:47Z","title":"MoL for LLMs: Dual-Loss Optimization to Enhance Domain Expertise While Preserving General Capabilities"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12043","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":"sha256:5b1795d294f07587a5b81ae099587a8b14c893945da94afee2b75649e6fcd967","target":"record","created_at":"2026-07-05T11:05:45Z","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":"9c13e3561321c32109d949b0b59f29adf95ef4e19e4d8ac0bb4d89d56379209a","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-17T15:12:47Z","title_canon_sha256":"2f24bfa8859bf9ca1ba1da9d02922bfe6fb99de5582c2c075d732c8a3a222e8e"},"schema_version":"1.0","source":{"id":"2505.12043","kind":"arxiv","version":2}},"canonical_sha256":"680612b9cbc192d340044691032baea991a71dd336d9ad2210dce4f4916f574c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"680612b9cbc192d340044691032baea991a71dd336d9ad2210dce4f4916f574c","first_computed_at":"2026-07-05T11:05:45.159783Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:05:45.159783Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"qmXq6+EWur6F8Vnjz+Qqlldv43L1p+0/cCyxYDEEmgiRGJmPQcnCwtMfyn0v2B8IjjELlR7wUSzpet2Y3YdcCg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:05:45.160247Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.12043","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5b1795d294f07587a5b81ae099587a8b14c893945da94afee2b75649e6fcd967","sha256:cd185b6884e4bf9b451d68ab09f62bbbf5d9e8a8087a894b1a337c48dc237710"],"state_sha256":"2380e8d9dee7484667e07aa0fa02210dfbe4956041ae89a5601cf9c350a914f1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Vam2idOT0MiB6XVSaU+l/hys3nQ2s5sg66GJyi50Kyd6/+dp5y/FZ6QtBuQgYw8n859AOBo1za3Y9PnR13NkCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T22:54:40.515999Z","bundle_sha256":"85d0284c8acc78e4fbcfbeba6f8b3b14c8f9b2f3907ed38e848fd0ec65f52910"}}