{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:T6VVSJDCEQZ6DFFPXSQ4SOKJSV","short_pith_number":"pith:T6VVSJDC","schema_version":"1.0","canonical_sha256":"9fab5924622433e194afbca1c939499577171e3641a73f23e3edeb4c10f9bd3c","source":{"kind":"arxiv","id":"2605.22205","version":1},"attestation_state":"computed","paper":{"title":"Skill Weaving: Efficient LLM Improvement via Modular Skillpacks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Guodong Du, Jiabo Zhang, Jing Li, Weijun Yao, Weiyang Guo, Yuan Zhou, Zesheng Shi, Zhuo Li","submitted_at":"2026-05-21T09:12:20Z","abstract_excerpt":"Large language models increasingly require specialization across diverse domains, yet existing approaches struggle to balance multi-domain capacities with strict memory and inference constraints. In this work, we introduce SkillWeave, a modular improvement framework that enables LLMs to specialize under fixed memory budgets. SkillWeave partitions full capabilities of a general-purpose model into skillpacks -- lightweight, domain-specific delta modules -- that reorganize and refine the model's internal knowledge. For efficient deployment, SkillWeave integrates SkillZip to compress skillpacks in"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2605.22205","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-21T09:12:20Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"37d01373f8f675a8eac3a8efd84fd1f5bea20bb6c7319b94005a41c37d3f6061","abstract_canon_sha256":"f91d6ce351bd74386f86226da552e1c7d28dfc696304937d933abab9e40ea972"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-22T01:04:32.143293Z","signature_b64":"t8QBPxagSTyOHzIOYFz2TPd0CdQyKSF2UvhjtbolvNX6ZWT/ds22PPAcg0rMGqJ62A7kuA4JlP5HD1o0Al5iDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9fab5924622433e194afbca1c939499577171e3641a73f23e3edeb4c10f9bd3c","last_reissued_at":"2026-05-22T01:04:32.142518Z","signature_status":"signed_v1","first_computed_at":"2026-05-22T01:04:32.142518Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Skill Weaving: Efficient LLM Improvement via Modular Skillpacks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Guodong Du, Jiabo Zhang, Jing Li, Weijun Yao, Weiyang Guo, Yuan Zhou, Zesheng Shi, Zhuo Li","submitted_at":"2026-05-21T09:12:20Z","abstract_excerpt":"Large language models increasingly require specialization across diverse domains, yet existing approaches struggle to balance multi-domain capacities with strict memory and inference constraints. In this work, we introduce SkillWeave, a modular improvement framework that enables LLMs to specialize under fixed memory budgets. SkillWeave partitions full capabilities of a general-purpose model into skillpacks -- lightweight, domain-specific delta modules -- that reorganize and refine the model's internal knowledge. For efficient deployment, SkillWeave integrates SkillZip to compress skillpacks in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.22205","kind":"arxiv","version":1},"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/2605.22205/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2605.22205","created_at":"2026-05-22T01:04:32.142635+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.22205v1","created_at":"2026-05-22T01:04:32.142635+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.22205","created_at":"2026-05-22T01:04:32.142635+00:00"},{"alias_kind":"pith_short_12","alias_value":"T6VVSJDCEQZ6","created_at":"2026-05-22T01:04:32.142635+00:00"},{"alias_kind":"pith_short_16","alias_value":"T6VVSJDCEQZ6DFFP","created_at":"2026-05-22T01:04:32.142635+00:00"},{"alias_kind":"pith_short_8","alias_value":"T6VVSJDC","created_at":"2026-05-22T01:04:32.142635+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/T6VVSJDCEQZ6DFFPXSQ4SOKJSV","json":"https://pith.science/pith/T6VVSJDCEQZ6DFFPXSQ4SOKJSV.json","graph_json":"https://pith.science/api/pith-number/T6VVSJDCEQZ6DFFPXSQ4SOKJSV/graph.json","events_json":"https://pith.science/api/pith-number/T6VVSJDCEQZ6DFFPXSQ4SOKJSV/events.json","paper":"https://pith.science/paper/T6VVSJDC"},"agent_actions":{"view_html":"https://pith.science/pith/T6VVSJDCEQZ6DFFPXSQ4SOKJSV","download_json":"https://pith.science/pith/T6VVSJDCEQZ6DFFPXSQ4SOKJSV.json","view_paper":"https://pith.science/paper/T6VVSJDC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.22205&json=true","fetch_graph":"https://pith.science/api/pith-number/T6VVSJDCEQZ6DFFPXSQ4SOKJSV/graph.json","fetch_events":"https://pith.science/api/pith-number/T6VVSJDCEQZ6DFFPXSQ4SOKJSV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/T6VVSJDCEQZ6DFFPXSQ4SOKJSV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/T6VVSJDCEQZ6DFFPXSQ4SOKJSV/action/storage_attestation","attest_author":"https://pith.science/pith/T6VVSJDCEQZ6DFFPXSQ4SOKJSV/action/author_attestation","sign_citation":"https://pith.science/pith/T6VVSJDCEQZ6DFFPXSQ4SOKJSV/action/citation_signature","submit_replication":"https://pith.science/pith/T6VVSJDCEQZ6DFFPXSQ4SOKJSV/action/replication_record"}},"created_at":"2026-05-22T01:04:32.142635+00:00","updated_at":"2026-05-22T01:04:32.142635+00:00"}