{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:7B4PWYPZBN4AGIBYVB64USDOTV","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":"fdb9164f67288954af4983c87fb382e3a2c28b52de5d4a5881b40ee6c6689e23","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-31T07:54:34Z","title_canon_sha256":"2de099aa7a5bdec6d2e3d45285adea6dcbd2039f0a1e5b7dd72694b270f992b5"},"schema_version":"1.0","source":{"id":"2502.00075","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.00075","created_at":"2026-07-05T10:08:02Z"},{"alias_kind":"arxiv_version","alias_value":"2502.00075v1","created_at":"2026-07-05T10:08:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.00075","created_at":"2026-07-05T10:08:02Z"},{"alias_kind":"pith_short_12","alias_value":"7B4PWYPZBN4A","created_at":"2026-07-05T10:08:02Z"},{"alias_kind":"pith_short_16","alias_value":"7B4PWYPZBN4AGIBY","created_at":"2026-07-05T10:08:02Z"},{"alias_kind":"pith_short_8","alias_value":"7B4PWYPZ","created_at":"2026-07-05T10:08:02Z"}],"graph_snapshots":[{"event_id":"sha256:7e189249e30d5e175ada2822ea7c28b11c119f8b7d3c899e2dce15f87a9f456a","target":"graph","created_at":"2026-07-05T10:08:02Z","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/2502.00075/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We present Branch-Train-Stitch (BTS), an efficient and flexible training algorithm for combining independently trained large language model (LLM) experts into a single, capable generalist model. Following Li et al., we start with a single seed language model which is branched into domain-specific (e.g., coding or math) experts with continual pretraining. BTS combines experts into a generalist model using lightweight stitch layers, which are inserted between frozen experts and the seed LLM, and trained on a small datamix of the expert domains. Stitch layers enable the seed LLM to integrate repr","authors_text":"Chloe Bi, Chris X. Cai, Emily Dinan, Jakob Foerster, Jeremy Fu, Mike Lewis, Prajjwal Bhargava, Punit Singh Koura, Qizhen Zhang, Ruan Silva, Sheng Shen, Suchin Gururangan","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-31T07:54:34Z","title":"BTS: Harmonizing Specialized Experts into a Generalist LLM"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.00075","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:f62c759dbb611a92b59b957c386462b6e984f0673d1d00f8c76bd25488d75ebe","target":"record","created_at":"2026-07-05T10:08:02Z","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":"fdb9164f67288954af4983c87fb382e3a2c28b52de5d4a5881b40ee6c6689e23","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-31T07:54:34Z","title_canon_sha256":"2de099aa7a5bdec6d2e3d45285adea6dcbd2039f0a1e5b7dd72694b270f992b5"},"schema_version":"1.0","source":{"id":"2502.00075","kind":"arxiv","version":1}},"canonical_sha256":"f878fb61f90b78032038a87dca486e9d634d0c3a9999a2a455e287c4563d8c7b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f878fb61f90b78032038a87dca486e9d634d0c3a9999a2a455e287c4563d8c7b","first_computed_at":"2026-07-05T10:08:02.765797Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:08:02.765797Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ICi+lSIvYbqIqegpuwDFKLGEQtcpWFjW6b8rKOXwJrp161qp6YaLaSAtCdxbxlzyQS5l5iinQzIMRMBsvIbGBg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:08:02.766291Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.00075","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f62c759dbb611a92b59b957c386462b6e984f0673d1d00f8c76bd25488d75ebe","sha256:7e189249e30d5e175ada2822ea7c28b11c119f8b7d3c899e2dce15f87a9f456a"],"state_sha256":"8182928a6a04b155769cc5d7f1dd626bf7cefb4eff652b339af6f6a38b3af0b8"}