{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:XGJRORL2NYDUJLO3SAC5MBOFJE","short_pith_number":"pith:XGJRORL2","schema_version":"1.0","canonical_sha256":"b99317457a6e0744addb9005d605c5492ed4a5ddd43e25885548efeec16f2279","source":{"kind":"arxiv","id":"2605.28424","version":1},"attestation_state":"computed","paper":{"title":"Skill0.5: Joint Skill Internalization and Utilization for Out-of-Distribution Generalization in Agentic Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chengcheng Han, Jianxiang Yu, Jiapeng Zhu, Qi Gu, Weining Qian, Xiang Li, Xunliang Cai, Yibo Zhao","submitted_at":"2026-05-27T12:54:33Z","abstract_excerpt":"Equipping large language models with explicit skills has emerged as a promising paradigm for enabling autonomous agents to solve complex tasks. Agent skills can be inherently divided into general skills for broad cognitive transfer and task-specific skills for dynamic execution. However, existing skill-based reinforcement learning (RL) methods typically force a rigid choice between full externalization, which incurs prohibitive context overhead, and full internalization, which risks overfitting and knowledge conflicts. To address this dilemma, we propose Skill0.5, a novel agentic RL framework "},"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":"2605.28424","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-27T12:54:33Z","cross_cats_sorted":[],"title_canon_sha256":"05ff34eea5646bf3525aea5fba8e75231b34e4afa87cb02b70bd93cbb7498433","abstract_canon_sha256":"82a4fb80f702c126ededc4505ab3650292ed55092b826673a0baacdac72433d0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-28T01:05:17.790911Z","signature_b64":"Gw1m+AAwCK2nh0Pwt+Oum2WEcZvlC3Vg69aaXStYBlZ3mJ2q2ckdUSRPUX58YlXiviOu7CMKakT+fdRzdkdmBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b99317457a6e0744addb9005d605c5492ed4a5ddd43e25885548efeec16f2279","last_reissued_at":"2026-05-28T01:05:17.790493Z","signature_status":"signed_v1","first_computed_at":"2026-05-28T01:05:17.790493Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Skill0.5: Joint Skill Internalization and Utilization for Out-of-Distribution Generalization in Agentic Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chengcheng Han, Jianxiang Yu, Jiapeng Zhu, Qi Gu, Weining Qian, Xiang Li, Xunliang Cai, Yibo Zhao","submitted_at":"2026-05-27T12:54:33Z","abstract_excerpt":"Equipping large language models with explicit skills has emerged as a promising paradigm for enabling autonomous agents to solve complex tasks. Agent skills can be inherently divided into general skills for broad cognitive transfer and task-specific skills for dynamic execution. However, existing skill-based reinforcement learning (RL) methods typically force a rigid choice between full externalization, which incurs prohibitive context overhead, and full internalization, which risks overfitting and knowledge conflicts. To address this dilemma, we propose Skill0.5, a novel agentic RL framework "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.28424","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/2605.28424/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":"2605.28424","created_at":"2026-05-28T01:05:17.790553+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.28424v1","created_at":"2026-05-28T01:05:17.790553+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.28424","created_at":"2026-05-28T01:05:17.790553+00:00"},{"alias_kind":"pith_short_12","alias_value":"XGJRORL2NYDU","created_at":"2026-05-28T01:05:17.790553+00:00"},{"alias_kind":"pith_short_16","alias_value":"XGJRORL2NYDUJLO3","created_at":"2026-05-28T01:05:17.790553+00:00"},{"alias_kind":"pith_short_8","alias_value":"XGJRORL2","created_at":"2026-05-28T01:05:17.790553+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/XGJRORL2NYDUJLO3SAC5MBOFJE","json":"https://pith.science/pith/XGJRORL2NYDUJLO3SAC5MBOFJE.json","graph_json":"https://pith.science/api/pith-number/XGJRORL2NYDUJLO3SAC5MBOFJE/graph.json","events_json":"https://pith.science/api/pith-number/XGJRORL2NYDUJLO3SAC5MBOFJE/events.json","paper":"https://pith.science/paper/XGJRORL2"},"agent_actions":{"view_html":"https://pith.science/pith/XGJRORL2NYDUJLO3SAC5MBOFJE","download_json":"https://pith.science/pith/XGJRORL2NYDUJLO3SAC5MBOFJE.json","view_paper":"https://pith.science/paper/XGJRORL2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.28424&json=true","fetch_graph":"https://pith.science/api/pith-number/XGJRORL2NYDUJLO3SAC5MBOFJE/graph.json","fetch_events":"https://pith.science/api/pith-number/XGJRORL2NYDUJLO3SAC5MBOFJE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XGJRORL2NYDUJLO3SAC5MBOFJE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XGJRORL2NYDUJLO3SAC5MBOFJE/action/storage_attestation","attest_author":"https://pith.science/pith/XGJRORL2NYDUJLO3SAC5MBOFJE/action/author_attestation","sign_citation":"https://pith.science/pith/XGJRORL2NYDUJLO3SAC5MBOFJE/action/citation_signature","submit_replication":"https://pith.science/pith/XGJRORL2NYDUJLO3SAC5MBOFJE/action/replication_record"}},"created_at":"2026-05-28T01:05:17.790553+00:00","updated_at":"2026-05-28T01:05:17.790553+00:00"}