{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:GSVQXGAUM2EW2LPJRP7BOYLZJF","short_pith_number":"pith:GSVQXGAU","schema_version":"1.0","canonical_sha256":"34ab0b981466896d2de98bfe176179495c03ce2664a00207adde57e87965377e","source":{"kind":"arxiv","id":"2410.14735","version":4},"attestation_state":"computed","paper":{"title":"Agent Skill Acquisition for Large Language Models via CycleQD","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.NE"],"primary_cat":"cs.CL","authors_text":"So Kuroki, Taishi Nakamura, Takuya Akiba, Yujin Tang","submitted_at":"2024-10-16T20:27:15Z","abstract_excerpt":"Training large language models to acquire specific skills remains a challenging endeavor. Conventional training approaches often struggle with data distribution imbalances and inadequacies in objective functions that do not align well with task-specific performance. To address these challenges, we introduce CycleQD, a novel approach that leverages the Quality Diversity framework through a cyclic adaptation of the algorithm, along with a model merging based crossover and an SVD-based mutation. In CycleQD, each task's performance metric is alternated as the quality measure while the others serve"},"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":"2410.14735","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-16T20:27:15Z","cross_cats_sorted":["cs.AI","cs.NE"],"title_canon_sha256":"e70ec7ebfbf1d542c4eb794940f77d9e77b0c01cc457f29ac4c0218f5621aeda","abstract_canon_sha256":"8d9ae14b92db3d1909791a6fb67e3af330a473259ccd3451a2b81d2ae786a275"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:15:20.737510Z","signature_b64":"e1CHjWnK4H95Jboh56yYh5TxyxFahQcfUYC09Hod4k7NcOGwpWgLfQjT4L83WFWfXUvR/oiRacjyjO2G3jmlCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"34ab0b981466896d2de98bfe176179495c03ce2664a00207adde57e87965377e","last_reissued_at":"2026-07-05T10:15:20.737019Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:15:20.737019Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Agent Skill Acquisition for Large Language Models via CycleQD","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.NE"],"primary_cat":"cs.CL","authors_text":"So Kuroki, Taishi Nakamura, Takuya Akiba, Yujin Tang","submitted_at":"2024-10-16T20:27:15Z","abstract_excerpt":"Training large language models to acquire specific skills remains a challenging endeavor. Conventional training approaches often struggle with data distribution imbalances and inadequacies in objective functions that do not align well with task-specific performance. To address these challenges, we introduce CycleQD, a novel approach that leverages the Quality Diversity framework through a cyclic adaptation of the algorithm, along with a model merging based crossover and an SVD-based mutation. In CycleQD, each task's performance metric is alternated as the quality measure while the others serve"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.14735","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/2410.14735/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":"2410.14735","created_at":"2026-07-05T10:15:20.737080+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.14735v4","created_at":"2026-07-05T10:15:20.737080+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.14735","created_at":"2026-07-05T10:15:20.737080+00:00"},{"alias_kind":"pith_short_12","alias_value":"GSVQXGAUM2EW","created_at":"2026-07-05T10:15:20.737080+00:00"},{"alias_kind":"pith_short_16","alias_value":"GSVQXGAUM2EW2LPJ","created_at":"2026-07-05T10:15:20.737080+00:00"},{"alias_kind":"pith_short_8","alias_value":"GSVQXGAU","created_at":"2026-07-05T10:15:20.737080+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.26200","citing_title":"Workflow Closure Is Not Scientific Closure in Auto-Research Systems","ref_index":59,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28898","citing_title":"PASTA: A Paraphrasing And Self-Training Approach for Knowledge Updating in LLMs","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.02721","citing_title":"GrandCode: Achieving Grandmaster Level in Competitive Programming via Agentic Reinforcement Learning","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11153","citing_title":"Decomposing Evolutionary Mixture-of-LoRA Architectures: The Routing Lever, the Lifecycle Penalty, and a Substrate-Conditional Boundary","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17503","citing_title":"SkillGraph: Self-Evolving Multi-Agent Collaboration with Multimodal Graph Topology","ref_index":22,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GSVQXGAUM2EW2LPJRP7BOYLZJF","json":"https://pith.science/pith/GSVQXGAUM2EW2LPJRP7BOYLZJF.json","graph_json":"https://pith.science/api/pith-number/GSVQXGAUM2EW2LPJRP7BOYLZJF/graph.json","events_json":"https://pith.science/api/pith-number/GSVQXGAUM2EW2LPJRP7BOYLZJF/events.json","paper":"https://pith.science/paper/GSVQXGAU"},"agent_actions":{"view_html":"https://pith.science/pith/GSVQXGAUM2EW2LPJRP7BOYLZJF","download_json":"https://pith.science/pith/GSVQXGAUM2EW2LPJRP7BOYLZJF.json","view_paper":"https://pith.science/paper/GSVQXGAU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.14735&json=true","fetch_graph":"https://pith.science/api/pith-number/GSVQXGAUM2EW2LPJRP7BOYLZJF/graph.json","fetch_events":"https://pith.science/api/pith-number/GSVQXGAUM2EW2LPJRP7BOYLZJF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GSVQXGAUM2EW2LPJRP7BOYLZJF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GSVQXGAUM2EW2LPJRP7BOYLZJF/action/storage_attestation","attest_author":"https://pith.science/pith/GSVQXGAUM2EW2LPJRP7BOYLZJF/action/author_attestation","sign_citation":"https://pith.science/pith/GSVQXGAUM2EW2LPJRP7BOYLZJF/action/citation_signature","submit_replication":"https://pith.science/pith/GSVQXGAUM2EW2LPJRP7BOYLZJF/action/replication_record"}},"created_at":"2026-07-05T10:15:20.737080+00:00","updated_at":"2026-07-05T10:15:20.737080+00:00"}