{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:HZXLF2PJ4VIUBNFJUV3VQGQYHB","short_pith_number":"pith:HZXLF2PJ","schema_version":"1.0","canonical_sha256":"3e6eb2e9e9e55140b4a9a577581a18384f058d1444993c53564f3ce17aa84d15","source":{"kind":"arxiv","id":"2512.09706","version":2},"attestation_state":"computed","paper":{"title":"Training One Model to Master Cross-Level Agentic Actions via Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Anji Liu, Kaichen He, Muyao Li, Yitao Liang, Zihao Wang","submitted_at":"2025-12-10T14:52:29Z","abstract_excerpt":"The paradigm of agentic AI is shifting from engineered complex workflows to post-training native models. However, existing agents are typically confined to static, predefined action spaces-such as exclusively using APIs, GUI events, or robotic commands. This rigidity limits their adaptability in dynamic environments where the optimal granularity of interaction varies contextually. To bridge this gap, we propose CrossHA, a unified agentic model that masters heterogeneous action spaces and autonomously selects the most effective interface for each step of a trajectory. We introduce a comprehensi"},"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":"2512.09706","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-12-10T14:52:29Z","cross_cats_sorted":[],"title_canon_sha256":"22feb77a5ad852a4ce12c94053af95edc924bbbb8ee1f6021b3aa043b0c686e3","abstract_canon_sha256":"4d0f7fd44e4da468efdb50939c84246fbbc609cba382debef8ae654fe4a6f3b8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T01:14:32.102476Z","signature_b64":"ky8Dww6WoLLDqgwvio3CvovUQtsIngrWWeYMs/cwWF4ZjT5EQnRLXLOtjlDZwH/GyzZ2zxc4tYzYomXD8L3wAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3e6eb2e9e9e55140b4a9a577581a18384f058d1444993c53564f3ce17aa84d15","last_reissued_at":"2026-06-05T01:14:32.101691Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T01:14:32.101691Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Training One Model to Master Cross-Level Agentic Actions via Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Anji Liu, Kaichen He, Muyao Li, Yitao Liang, Zihao Wang","submitted_at":"2025-12-10T14:52:29Z","abstract_excerpt":"The paradigm of agentic AI is shifting from engineered complex workflows to post-training native models. However, existing agents are typically confined to static, predefined action spaces-such as exclusively using APIs, GUI events, or robotic commands. This rigidity limits their adaptability in dynamic environments where the optimal granularity of interaction varies contextually. To bridge this gap, we propose CrossHA, a unified agentic model that masters heterogeneous action spaces and autonomously selects the most effective interface for each step of a trajectory. We introduce a comprehensi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2512.09706","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/2512.09706/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":"2512.09706","created_at":"2026-06-05T01:14:32.101799+00:00"},{"alias_kind":"arxiv_version","alias_value":"2512.09706v2","created_at":"2026-06-05T01:14:32.101799+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.09706","created_at":"2026-06-05T01:14:32.101799+00:00"},{"alias_kind":"pith_short_12","alias_value":"HZXLF2PJ4VIU","created_at":"2026-06-05T01:14:32.101799+00:00"},{"alias_kind":"pith_short_16","alias_value":"HZXLF2PJ4VIUBNFJ","created_at":"2026-06-05T01:14:32.101799+00:00"},{"alias_kind":"pith_short_8","alias_value":"HZXLF2PJ","created_at":"2026-06-05T01:14:32.101799+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/HZXLF2PJ4VIUBNFJUV3VQGQYHB","json":"https://pith.science/pith/HZXLF2PJ4VIUBNFJUV3VQGQYHB.json","graph_json":"https://pith.science/api/pith-number/HZXLF2PJ4VIUBNFJUV3VQGQYHB/graph.json","events_json":"https://pith.science/api/pith-number/HZXLF2PJ4VIUBNFJUV3VQGQYHB/events.json","paper":"https://pith.science/paper/HZXLF2PJ"},"agent_actions":{"view_html":"https://pith.science/pith/HZXLF2PJ4VIUBNFJUV3VQGQYHB","download_json":"https://pith.science/pith/HZXLF2PJ4VIUBNFJUV3VQGQYHB.json","view_paper":"https://pith.science/paper/HZXLF2PJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2512.09706&json=true","fetch_graph":"https://pith.science/api/pith-number/HZXLF2PJ4VIUBNFJUV3VQGQYHB/graph.json","fetch_events":"https://pith.science/api/pith-number/HZXLF2PJ4VIUBNFJUV3VQGQYHB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HZXLF2PJ4VIUBNFJUV3VQGQYHB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HZXLF2PJ4VIUBNFJUV3VQGQYHB/action/storage_attestation","attest_author":"https://pith.science/pith/HZXLF2PJ4VIUBNFJUV3VQGQYHB/action/author_attestation","sign_citation":"https://pith.science/pith/HZXLF2PJ4VIUBNFJUV3VQGQYHB/action/citation_signature","submit_replication":"https://pith.science/pith/HZXLF2PJ4VIUBNFJUV3VQGQYHB/action/replication_record"}},"created_at":"2026-06-05T01:14:32.101799+00:00","updated_at":"2026-06-05T01:14:32.101799+00:00"}