{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:KGVPGJVIXDLASUQZ6T5KFZX5C4","short_pith_number":"pith:KGVPGJVI","schema_version":"1.0","canonical_sha256":"51aaf326a8b8d6095219f4faa2e6fd171ef6d911b4a1e1e300d53a7abdb9f32b","source":{"kind":"arxiv","id":"2508.18040","version":1},"attestation_state":"computed","paper":{"title":"PerPilot: Personalizing VLM-based Mobile Agents via Memory and Exploration","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chenxu Wang, Hao Li, Jinzhuo Liu, Kun Shao, Qiaosheng Zhang, Ruiqi Song, Shuyue Hu, Siyue Ren, Xin Wang, Ya Zeng, Yihang Chen, Zhen Wang, Zhiyao Cui","submitted_at":"2025-08-25T13:57:02Z","abstract_excerpt":"Vision language model (VLM)-based mobile agents show great potential for assisting users in performing instruction-driven tasks. However, these agents typically struggle with personalized instructions -- those containing ambiguous, user-specific context -- a challenge that has been largely overlooked in previous research. In this paper, we define personalized instructions and introduce PerInstruct, a novel human-annotated dataset covering diverse personalized instructions across various mobile scenarios. Furthermore, given the limited personalization capabilities of existing mobile agents, we "},"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":"2508.18040","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-08-25T13:57:02Z","cross_cats_sorted":[],"title_canon_sha256":"22c3938eb1cb145b9e29a69bbbf0c9eaa266d582c7243ee2a7cc65501b17b9a5","abstract_canon_sha256":"49bb6b936d8d035de15909dbf8adfafa84735ec4792f304c502232b646b7995d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:58:52.489640Z","signature_b64":"ccehhOTVzjf6mjHbdp2ZawLOPCaN33hZ8VvHCgZu1iS621SO2AwAuOVfLpzoK6YJA8TOa3yVLAAsZUGD7jGLCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"51aaf326a8b8d6095219f4faa2e6fd171ef6d911b4a1e1e300d53a7abdb9f32b","last_reissued_at":"2026-07-05T11:58:52.489130Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:58:52.489130Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PerPilot: Personalizing VLM-based Mobile Agents via Memory and Exploration","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chenxu Wang, Hao Li, Jinzhuo Liu, Kun Shao, Qiaosheng Zhang, Ruiqi Song, Shuyue Hu, Siyue Ren, Xin Wang, Ya Zeng, Yihang Chen, Zhen Wang, Zhiyao Cui","submitted_at":"2025-08-25T13:57:02Z","abstract_excerpt":"Vision language model (VLM)-based mobile agents show great potential for assisting users in performing instruction-driven tasks. However, these agents typically struggle with personalized instructions -- those containing ambiguous, user-specific context -- a challenge that has been largely overlooked in previous research. In this paper, we define personalized instructions and introduce PerInstruct, a novel human-annotated dataset covering diverse personalized instructions across various mobile scenarios. Furthermore, given the limited personalization capabilities of existing mobile agents, we "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.18040","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/2508.18040/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":"2508.18040","created_at":"2026-07-05T11:58:52.489187+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.18040v1","created_at":"2026-07-05T11:58:52.489187+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.18040","created_at":"2026-07-05T11:58:52.489187+00:00"},{"alias_kind":"pith_short_12","alias_value":"KGVPGJVIXDLA","created_at":"2026-07-05T11:58:52.489187+00:00"},{"alias_kind":"pith_short_16","alias_value":"KGVPGJVIXDLASUQZ","created_at":"2026-07-05T11:58:52.489187+00:00"},{"alias_kind":"pith_short_8","alias_value":"KGVPGJVI","created_at":"2026-07-05T11:58:52.489187+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2509.07553","citing_title":"VeriOS: Query-Driven Proactive Human-Agent-GUI Interaction for Trustworthy OS Agents","ref_index":49,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KGVPGJVIXDLASUQZ6T5KFZX5C4","json":"https://pith.science/pith/KGVPGJVIXDLASUQZ6T5KFZX5C4.json","graph_json":"https://pith.science/api/pith-number/KGVPGJVIXDLASUQZ6T5KFZX5C4/graph.json","events_json":"https://pith.science/api/pith-number/KGVPGJVIXDLASUQZ6T5KFZX5C4/events.json","paper":"https://pith.science/paper/KGVPGJVI"},"agent_actions":{"view_html":"https://pith.science/pith/KGVPGJVIXDLASUQZ6T5KFZX5C4","download_json":"https://pith.science/pith/KGVPGJVIXDLASUQZ6T5KFZX5C4.json","view_paper":"https://pith.science/paper/KGVPGJVI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.18040&json=true","fetch_graph":"https://pith.science/api/pith-number/KGVPGJVIXDLASUQZ6T5KFZX5C4/graph.json","fetch_events":"https://pith.science/api/pith-number/KGVPGJVIXDLASUQZ6T5KFZX5C4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KGVPGJVIXDLASUQZ6T5KFZX5C4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KGVPGJVIXDLASUQZ6T5KFZX5C4/action/storage_attestation","attest_author":"https://pith.science/pith/KGVPGJVIXDLASUQZ6T5KFZX5C4/action/author_attestation","sign_citation":"https://pith.science/pith/KGVPGJVIXDLASUQZ6T5KFZX5C4/action/citation_signature","submit_replication":"https://pith.science/pith/KGVPGJVIXDLASUQZ6T5KFZX5C4/action/replication_record"}},"created_at":"2026-07-05T11:58:52.489187+00:00","updated_at":"2026-07-05T11:58:52.489187+00:00"}