{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:ERPANQULII74QF7NLANSX4ZAAQ","short_pith_number":"pith:ERPANQUL","schema_version":"1.0","canonical_sha256":"245e06c28b423fc817ed581b2bf32004024fd19b65b4fea22cf19cc7685ef25b","source":{"kind":"arxiv","id":"2601.22758","version":2},"attestation_state":"computed","paper":{"title":"AutoRefine: Compiling Trajectories into Validated Typed Agent Artifacts","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Junfu Chen, Liangyu Li, Libin Qiu, Shuo Tang, Weizhi Huang, Wenkai Qiu, Xiaobo Xue, Yuhang Ye, Zhirong Gao","submitted_at":"2026-01-30T09:33:30Z","abstract_excerpt":"Large language model agents repeatedly encounter related tasks, yet systems that learn from trajectories commit every lesson to one predefined artifact form. A local constraint, a reusable procedure, and a delegated objective require different amounts of runtime ownership, so one form either under-specifies the correction or wraps it in execution machinery it does not need. We present AutoRefine, which treats trajectory learning as typed artifact compilation. It contrasts failed against successful trajectories to derive a type-neutral, evidence-linked intervention specification, then compiles "},"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":"2601.22758","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-01-30T09:33:30Z","cross_cats_sorted":[],"title_canon_sha256":"3c7f03d082546a200c594da95468db83aedd049bcbed6e927ebfff9d18ce8710","abstract_canon_sha256":"e926c985fc955ede807578acb49211c0be19292dd5d904b47c14b469ce220c0d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T02:18:48.373917Z","signature_b64":"QRkJVzvLm673gb06Nct6B9J91xBueQ9SeeJFp1UXJ5d8aAQ2F1KeXPK5Ll0tbR8uBVgUFzCX73+Oq9BwL+gKCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"245e06c28b423fc817ed581b2bf32004024fd19b65b4fea22cf19cc7685ef25b","last_reissued_at":"2026-08-11T02:18:48.370824Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T02:18:48.370824Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AutoRefine: Compiling Trajectories into Validated Typed Agent Artifacts","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Junfu Chen, Liangyu Li, Libin Qiu, Shuo Tang, Weizhi Huang, Wenkai Qiu, Xiaobo Xue, Yuhang Ye, Zhirong Gao","submitted_at":"2026-01-30T09:33:30Z","abstract_excerpt":"Large language model agents repeatedly encounter related tasks, yet systems that learn from trajectories commit every lesson to one predefined artifact form. A local constraint, a reusable procedure, and a delegated objective require different amounts of runtime ownership, so one form either under-specifies the correction or wraps it in execution machinery it does not need. We present AutoRefine, which treats trajectory learning as typed artifact compilation. It contrasts failed against successful trajectories to derive a type-neutral, evidence-linked intervention specification, then compiles "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.22758","kind":"arxiv","version":2},"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/2601.22758/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":"2601.22758","created_at":"2026-08-11T02:18:48.372550+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.22758v2","created_at":"2026-08-11T02:18:48.372550+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.22758","created_at":"2026-08-11T02:18:48.372550+00:00"},{"alias_kind":"pith_short_12","alias_value":"ERPANQULII74","created_at":"2026-08-11T02:18:48.372550+00:00"},{"alias_kind":"pith_short_16","alias_value":"ERPANQULII74QF7N","created_at":"2026-08-11T02:18:48.372550+00:00"},{"alias_kind":"pith_short_8","alias_value":"ERPANQUL","created_at":"2026-08-11T02:18:48.372550+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":13,"internal_anchor_count":13,"sample":[{"citing_arxiv_id":"2605.09192","citing_title":"Evidence Over Plans: Online Trajectory Verification for Skill Distillation","ref_index":15,"is_internal_anchor":true},{"citing_arxiv_id":"2605.07358","citing_title":"A Comprehensive Survey on Agent Skills: Taxonomy, Techniques, and Applications","ref_index":76,"is_internal_anchor":true},{"citing_arxiv_id":"2605.23904","citing_title":"SkillOpt: Executive Strategy for Self-Evolving Agent Skills","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2606.29315","citing_title":"Hierarchical Experimentalist Agents","ref_index":3,"is_internal_anchor":true},{"citing_arxiv_id":"2605.23904","citing_title":"SkillOpt: Executive Strategy for Self-Evolving Agent Skills","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2605.23899","citing_title":"From Raw Experience to Skill Consumption: A Systematic Study of Model-Generated Agent Skills","ref_index":4,"is_internal_anchor":true},{"citing_arxiv_id":"2605.07358","citing_title":"A Comprehensive Survey on Agent Skills: Taxonomy, Techniques, and Applications","ref_index":78,"is_internal_anchor":true},{"citing_arxiv_id":"2605.14477","citing_title":"Test-Time Learning with an Evolving Library","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2604.27660","citing_title":"From Context to Skills: Can Language Models Learn from Context Skillfully?","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"2605.09192","citing_title":"Evidence Over Plans: Online Trajectory Verification for Skill Distillation","ref_index":12,"is_internal_anchor":true},{"citing_arxiv_id":"2605.07358","citing_title":"A Comprehensive Survey on Agent Skills: Taxonomy, Techniques, and Applications","ref_index":78,"is_internal_anchor":true},{"citing_arxiv_id":"2605.02913","citing_title":"Generate, Filter, Control, Replay: A Comprehensive Survey of Rollout Strategies for LLM Reinforcement Learning","ref_index":94,"is_internal_anchor":true},{"citing_arxiv_id":"2604.15097","citing_title":"From Procedural Skills to Strategy Genes: Towards Experience-Driven Test-Time Evolution","ref_index":14,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ERPANQULII74QF7NLANSX4ZAAQ","json":"https://pith.science/pith/ERPANQULII74QF7NLANSX4ZAAQ.json","graph_json":"https://pith.science/api/pith-number/ERPANQULII74QF7NLANSX4ZAAQ/graph.json","events_json":"https://pith.science/api/pith-number/ERPANQULII74QF7NLANSX4ZAAQ/events.json","paper":"https://pith.science/paper/ERPANQUL"},"agent_actions":{"view_html":"https://pith.science/pith/ERPANQULII74QF7NLANSX4ZAAQ","download_json":"https://pith.science/pith/ERPANQULII74QF7NLANSX4ZAAQ.json","view_paper":"https://pith.science/paper/ERPANQUL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.22758&json=true","fetch_graph":"https://pith.science/api/pith-number/ERPANQULII74QF7NLANSX4ZAAQ/graph.json","fetch_events":"https://pith.science/api/pith-number/ERPANQULII74QF7NLANSX4ZAAQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ERPANQULII74QF7NLANSX4ZAAQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ERPANQULII74QF7NLANSX4ZAAQ/action/storage_attestation","attest_author":"https://pith.science/pith/ERPANQULII74QF7NLANSX4ZAAQ/action/author_attestation","sign_citation":"https://pith.science/pith/ERPANQULII74QF7NLANSX4ZAAQ/action/citation_signature","submit_replication":"https://pith.science/pith/ERPANQULII74QF7NLANSX4ZAAQ/action/replication_record"}},"created_at":"2026-08-11T02:18:48.372550+00:00","updated_at":"2026-08-11T02:18:48.372550+00:00"}