{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:V3KE6ABM7MU2WKVB6XSBV4T4ZF","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":"9e469e9c2647a03f497d4d8b626ab89f9db2cb2f07118fcb859a085289b93344","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-18T08:59:08Z","title_canon_sha256":"890d7396d492d64ee7f0a2cc1597605b350fc8e72e0bbfe89891dd743f8461c4"},"schema_version":"1.0","source":{"id":"2605.18083","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.18083","created_at":"2026-05-20T00:05:15Z"},{"alias_kind":"arxiv_version","alias_value":"2605.18083v1","created_at":"2026-05-20T00:05:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.18083","created_at":"2026-05-20T00:05:15Z"},{"alias_kind":"pith_short_12","alias_value":"V3KE6ABM7MU2","created_at":"2026-05-20T00:05:15Z"},{"alias_kind":"pith_short_16","alias_value":"V3KE6ABM7MU2WKVB","created_at":"2026-05-20T00:05:15Z"},{"alias_kind":"pith_short_8","alias_value":"V3KE6ABM","created_at":"2026-05-20T00:05:15Z"}],"graph_snapshots":[{"event_id":"sha256:a7a4a256afb27d597f1031de6ae3689c76a641955e71f74cd8eabe4702ab63ba","target":"graph","created_at":"2026-05-20T00:05:15Z","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":[{"findings_count":0,"name":"claim_evidence","ran_at":"2026-05-19T23:41:59.222701Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-19T23:33:35.449389Z","status":"skipped","version":"1.0.0"}],"endpoint":"/pith/2605.18083/integrity.json","findings":[],"snapshot_sha256":"dbe1af9199807909ac8aa1bd8786b18df9ea03f5412ba6f4c11905915638a60f","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Expanding Large Language Models~(LLMs) to new languages is a costly endeavor, demanding extensive Continued Pre-Training~(CPT) and data-intensive alignment. While recent data-free merging techniques attempt to bypass alignment by fusing a multilingual CPT-enhanced model with its instruct counterpart, they are plagued by a critical trade-off: mitigating parameter conflicts to preserve original abilities inevitably dilutes new language acquisition, and vice-versa. To resolve this conflict, we introduce \\method, which upcycles a dense model into a Mixture-of-Experts~(MoE) architecture, allocating","authors_text":"Baosong Yang, Hao-Ran Wei, Hao Zhou, Jiajun Chen, Linjuan Wu, Shuaijie She, Shujian Huang, Tianhao Li, Zhijun Wang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-18T08:59:08Z","title":"A Data-Efficient Path to Multilingual LLMs: Language Expansion via Post-training PARAM$\\Delta$ Integration into Upcycled MoE"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.18083","kind":"arxiv","version":1},"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:f772af07eb5b7401a6907f2fd9376a1c826522ce91f920787ead5cf225d93c20","target":"record","created_at":"2026-05-20T00:05:15Z","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":"9e469e9c2647a03f497d4d8b626ab89f9db2cb2f07118fcb859a085289b93344","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-18T08:59:08Z","title_canon_sha256":"890d7396d492d64ee7f0a2cc1597605b350fc8e72e0bbfe89891dd743f8461c4"},"schema_version":"1.0","source":{"id":"2605.18083","kind":"arxiv","version":1}},"canonical_sha256":"aed44f002cfb29ab2aa1f5e41af27cc95c377a70514ea34497ad7223ad3b79f4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"aed44f002cfb29ab2aa1f5e41af27cc95c377a70514ea34497ad7223ad3b79f4","first_computed_at":"2026-05-20T00:05:15.189583Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:05:15.189583Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"H+KOKDVKWA928JkLhPdPeLOquttixSE4/lMrD9uSzPSRMqmF1nUvCXC3+5rUKD96zCy2u43VDX8ynzYRni5QBw==","signature_status":"signed_v1","signed_at":"2026-05-20T00:05:15.190498Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.18083","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f772af07eb5b7401a6907f2fd9376a1c826522ce91f920787ead5cf225d93c20","sha256:a7a4a256afb27d597f1031de6ae3689c76a641955e71f74cd8eabe4702ab63ba"],"state_sha256":"b6aaa03d454cfb57445559dd75b8f47324ccfd922c8bd48131318ee7ef6f1056"}