{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:NZ7RDKPLYL6MDS5VAOAPOZKSBW","short_pith_number":"pith:NZ7RDKPL","schema_version":"1.0","canonical_sha256":"6e7f11a9ebc2fcc1cbb50380f765520dab39445ab91e48121fe95061d2df6c02","source":{"kind":"arxiv","id":"2601.20209","version":2},"attestation_state":"computed","paper":{"title":"Spark: Strategic Policy-Aware Exploration via Dynamic Branching for Long-Horizon Agentic Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Changpeng Yang, Jianhua Tao, Jinyang Wu, Shuai Zhang, Shuo Yang, Yuhao Shen, Zhengqi Wen","submitted_at":"2026-01-28T03:15:34Z","abstract_excerpt":"Reinforcement learning has empowered large language models to act as intelligent agents, yet training them for long-horizon tasks remains challenging due to the scarcity of high-quality trajectories, especially under limited resources. Existing methods typically scale up rollout sizes and indiscriminately allocate computational resources among intermediate steps. Such attempts inherently waste substantial computation budget on trivial steps while failing to guarantee sample quality. To address this, we propose \\textbf{Spark} (\\textbf{S}trategic \\textbf{P}olicy-\\textbf{A}ware explo\\textbf{R}ati"},"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":"2601.20209","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-01-28T03:15:34Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"94568ca2741144c459827453d3052d2e8b41ef97f7d643d18aee4d7378588089","abstract_canon_sha256":"402007544d82d119fcebbe7ed40b374d1834b4e89a08781abcfea65634877943"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T02:12:46.189000Z","signature_b64":"c5BKpHWpq8undotOO49umB1L1O7w1FrDrSusbFzQjOAEoQAiFkvNEVsc1VbeCFgvHHGmNYMZx9Lcgzyb9k38CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6e7f11a9ebc2fcc1cbb50380f765520dab39445ab91e48121fe95061d2df6c02","last_reissued_at":"2026-06-23T02:12:46.188515Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T02:12:46.188515Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Spark: Strategic Policy-Aware Exploration via Dynamic Branching for Long-Horizon Agentic Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Changpeng Yang, Jianhua Tao, Jinyang Wu, Shuai Zhang, Shuo Yang, Yuhao Shen, Zhengqi Wen","submitted_at":"2026-01-28T03:15:34Z","abstract_excerpt":"Reinforcement learning has empowered large language models to act as intelligent agents, yet training them for long-horizon tasks remains challenging due to the scarcity of high-quality trajectories, especially under limited resources. Existing methods typically scale up rollout sizes and indiscriminately allocate computational resources among intermediate steps. Such attempts inherently waste substantial computation budget on trivial steps while failing to guarantee sample quality. To address this, we propose \\textbf{Spark} (\\textbf{S}trategic \\textbf{P}olicy-\\textbf{A}ware explo\\textbf{R}ati"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.20209","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/2601.20209/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":"2601.20209","created_at":"2026-06-23T02:12:46.188572+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.20209v2","created_at":"2026-06-23T02:12:46.188572+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.20209","created_at":"2026-06-23T02:12:46.188572+00:00"},{"alias_kind":"pith_short_12","alias_value":"NZ7RDKPLYL6M","created_at":"2026-06-23T02:12:46.188572+00:00"},{"alias_kind":"pith_short_16","alias_value":"NZ7RDKPLYL6MDS5V","created_at":"2026-06-23T02:12:46.188572+00:00"},{"alias_kind":"pith_short_8","alias_value":"NZ7RDKPL","created_at":"2026-06-23T02:12:46.188572+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":3,"sample":[{"citing_arxiv_id":"2605.22177","citing_title":"Maestro: Reinforcement Learning to Orchestrate Hierarchical Model-Skill Ensembles","ref_index":61,"is_internal_anchor":true},{"citing_arxiv_id":"2604.10693","citing_title":"FACT-E: Causality-Inspired Evaluation for Trustworthy Chain-of-Thought Reasoning","ref_index":40,"is_internal_anchor":true},{"citing_arxiv_id":"2604.09000","citing_title":"StreamMeCo: Long-Term Agent Memory Compression for Efficient Streaming Video Understanding","ref_index":5,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NZ7RDKPLYL6MDS5VAOAPOZKSBW","json":"https://pith.science/pith/NZ7RDKPLYL6MDS5VAOAPOZKSBW.json","graph_json":"https://pith.science/api/pith-number/NZ7RDKPLYL6MDS5VAOAPOZKSBW/graph.json","events_json":"https://pith.science/api/pith-number/NZ7RDKPLYL6MDS5VAOAPOZKSBW/events.json","paper":"https://pith.science/paper/NZ7RDKPL"},"agent_actions":{"view_html":"https://pith.science/pith/NZ7RDKPLYL6MDS5VAOAPOZKSBW","download_json":"https://pith.science/pith/NZ7RDKPLYL6MDS5VAOAPOZKSBW.json","view_paper":"https://pith.science/paper/NZ7RDKPL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.20209&json=true","fetch_graph":"https://pith.science/api/pith-number/NZ7RDKPLYL6MDS5VAOAPOZKSBW/graph.json","fetch_events":"https://pith.science/api/pith-number/NZ7RDKPLYL6MDS5VAOAPOZKSBW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NZ7RDKPLYL6MDS5VAOAPOZKSBW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NZ7RDKPLYL6MDS5VAOAPOZKSBW/action/storage_attestation","attest_author":"https://pith.science/pith/NZ7RDKPLYL6MDS5VAOAPOZKSBW/action/author_attestation","sign_citation":"https://pith.science/pith/NZ7RDKPLYL6MDS5VAOAPOZKSBW/action/citation_signature","submit_replication":"https://pith.science/pith/NZ7RDKPLYL6MDS5VAOAPOZKSBW/action/replication_record"}},"created_at":"2026-06-23T02:12:46.188572+00:00","updated_at":"2026-06-23T02:12:46.188572+00:00"}