{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XDPC4DDSQ7HJ3BZXV5APR2RWKZ","short_pith_number":"pith:XDPC4DDS","schema_version":"1.0","canonical_sha256":"b8de2e0c7287ce9d8737af40f8ea365674ba08678f8a7153cd7bf6043e45f5f2","source":{"kind":"arxiv","id":"2506.08972","version":1},"attestation_state":"computed","paper":{"title":"Atomic-to-Compositional Generalization for Mobile Agents with A New Benchmark and Scheduling System","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ming Zhou, Pengzhou Cheng, Tingjia Miao, Yuan Guo, Zheng Wu, Zhuosheng Zhang","submitted_at":"2025-06-10T16:45:29Z","abstract_excerpt":"Autonomous agents powered by multimodal large language models have been developed to facilitate task execution on mobile devices. However, prior work has predominantly focused on atomic tasks -- such as shot-chain execution tasks and single-screen grounding tasks -- while overlooking the generalization to compositional tasks, which are indispensable for real-world applications. This work introduces UI-NEXUS, a comprehensive benchmark designed to evaluate mobile agents on three categories of compositional operations: Simple Concatenation, Context Transition, and Deep Dive. UI-NEXUS supports int"},"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":"2506.08972","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-10T16:45:29Z","cross_cats_sorted":[],"title_canon_sha256":"6801852a3101ab6968bd46d83a7ffbfae6cad8a1767aac827362de269d49d3d9","abstract_canon_sha256":"39924eaf2e49677108dec38de82cd529419d9c60470e724c022a157884175075"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:19:16.258665Z","signature_b64":"gYfWfhEc3zqQSx45qcirW2iLpt3A30YFgpnqKUEjetrbUn8NDMSUejZeE17mXYfm/pgDcwZw/4V1C5JCjpU4CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b8de2e0c7287ce9d8737af40f8ea365674ba08678f8a7153cd7bf6043e45f5f2","last_reissued_at":"2026-07-05T11:19:16.258235Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:19:16.258235Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Atomic-to-Compositional Generalization for Mobile Agents with A New Benchmark and Scheduling System","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ming Zhou, Pengzhou Cheng, Tingjia Miao, Yuan Guo, Zheng Wu, Zhuosheng Zhang","submitted_at":"2025-06-10T16:45:29Z","abstract_excerpt":"Autonomous agents powered by multimodal large language models have been developed to facilitate task execution on mobile devices. However, prior work has predominantly focused on atomic tasks -- such as shot-chain execution tasks and single-screen grounding tasks -- while overlooking the generalization to compositional tasks, which are indispensable for real-world applications. This work introduces UI-NEXUS, a comprehensive benchmark designed to evaluate mobile agents on three categories of compositional operations: Simple Concatenation, Context Transition, and Deep Dive. UI-NEXUS supports int"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.08972","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/2506.08972/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":"2506.08972","created_at":"2026-07-05T11:19:16.258294+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.08972v1","created_at":"2026-07-05T11:19:16.258294+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.08972","created_at":"2026-07-05T11:19:16.258294+00:00"},{"alias_kind":"pith_short_12","alias_value":"XDPC4DDSQ7HJ","created_at":"2026-07-05T11:19:16.258294+00:00"},{"alias_kind":"pith_short_16","alias_value":"XDPC4DDSQ7HJ3BZX","created_at":"2026-07-05T11:19:16.258294+00:00"},{"alias_kind":"pith_short_8","alias_value":"XDPC4DDS","created_at":"2026-07-05T11:19:16.258294+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/XDPC4DDSQ7HJ3BZXV5APR2RWKZ","json":"https://pith.science/pith/XDPC4DDSQ7HJ3BZXV5APR2RWKZ.json","graph_json":"https://pith.science/api/pith-number/XDPC4DDSQ7HJ3BZXV5APR2RWKZ/graph.json","events_json":"https://pith.science/api/pith-number/XDPC4DDSQ7HJ3BZXV5APR2RWKZ/events.json","paper":"https://pith.science/paper/XDPC4DDS"},"agent_actions":{"view_html":"https://pith.science/pith/XDPC4DDSQ7HJ3BZXV5APR2RWKZ","download_json":"https://pith.science/pith/XDPC4DDSQ7HJ3BZXV5APR2RWKZ.json","view_paper":"https://pith.science/paper/XDPC4DDS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.08972&json=true","fetch_graph":"https://pith.science/api/pith-number/XDPC4DDSQ7HJ3BZXV5APR2RWKZ/graph.json","fetch_events":"https://pith.science/api/pith-number/XDPC4DDSQ7HJ3BZXV5APR2RWKZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XDPC4DDSQ7HJ3BZXV5APR2RWKZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XDPC4DDSQ7HJ3BZXV5APR2RWKZ/action/storage_attestation","attest_author":"https://pith.science/pith/XDPC4DDSQ7HJ3BZXV5APR2RWKZ/action/author_attestation","sign_citation":"https://pith.science/pith/XDPC4DDSQ7HJ3BZXV5APR2RWKZ/action/citation_signature","submit_replication":"https://pith.science/pith/XDPC4DDSQ7HJ3BZXV5APR2RWKZ/action/replication_record"}},"created_at":"2026-07-05T11:19:16.258294+00:00","updated_at":"2026-07-05T11:19:16.258294+00:00"}