{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:SWB3SLGEOZBK2N4A3B2BKA6H5Z","short_pith_number":"pith:SWB3SLGE","schema_version":"1.0","canonical_sha256":"9583b92cc47642ad3780d8741503c7ee7b3e53c0b6cc349bafab65c25995ffcf","source":{"kind":"arxiv","id":"2311.07562","version":1},"attestation_state":"computed","paper":{"title":"GPT-4V in Wonderland: Large Multimodal Models for Zero-Shot Smartphone GUI Navigation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"An Yan, Jianfeng Gao, Jianfeng Wang, Jianwei Yang, Julian McAuley, Kevin Lin, Lijuan Wang, Linjie Li, Wanrong Zhu, Yiwu Zhong, Zhengyuan Yang, Zicheng Liu","submitted_at":"2023-11-13T18:53:37Z","abstract_excerpt":"We present MM-Navigator, a GPT-4V-based agent for the smartphone graphical user interface (GUI) navigation task. MM-Navigator can interact with a smartphone screen as human users, and determine subsequent actions to fulfill given instructions. Our findings demonstrate that large multimodal models (LMMs), specifically GPT-4V, excel in zero-shot GUI navigation through its advanced screen interpretation, action reasoning, and precise action localization capabilities. We first benchmark MM-Navigator on our collected iOS screen dataset. According to human assessments, the system exhibited a 91\\% ac"},"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":"2311.07562","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-11-13T18:53:37Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a4d4f19a887cbad7c99a8613bb54bf620ada14afc4928d6f3bb01ed591098282","abstract_canon_sha256":"5322677ed9934a592033fc71a9739494f276cba6ff69a11c14cfcdceadf94f22"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:12:13.939751Z","signature_b64":"PuKi4ZtZKXREiGrbGyvlNzcsuXgAYJSR+k43kBL1ClYQLGsz7SZAzIXt2rgVcqn0nczj0rRBOTjqIpcPLcqHCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9583b92cc47642ad3780d8741503c7ee7b3e53c0b6cc349bafab65c25995ffcf","last_reissued_at":"2026-07-05T07:12:13.939268Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:12:13.939268Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"GPT-4V in Wonderland: Large Multimodal Models for Zero-Shot Smartphone GUI Navigation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"An Yan, Jianfeng Gao, Jianfeng Wang, Jianwei Yang, Julian McAuley, Kevin Lin, Lijuan Wang, Linjie Li, Wanrong Zhu, Yiwu Zhong, Zhengyuan Yang, Zicheng Liu","submitted_at":"2023-11-13T18:53:37Z","abstract_excerpt":"We present MM-Navigator, a GPT-4V-based agent for the smartphone graphical user interface (GUI) navigation task. MM-Navigator can interact with a smartphone screen as human users, and determine subsequent actions to fulfill given instructions. Our findings demonstrate that large multimodal models (LMMs), specifically GPT-4V, excel in zero-shot GUI navigation through its advanced screen interpretation, action reasoning, and precise action localization capabilities. We first benchmark MM-Navigator on our collected iOS screen dataset. According to human assessments, the system exhibited a 91\\% ac"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.07562","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/2311.07562/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":"2311.07562","created_at":"2026-07-05T07:12:13.939331+00:00"},{"alias_kind":"arxiv_version","alias_value":"2311.07562v1","created_at":"2026-07-05T07:12:13.939331+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.07562","created_at":"2026-07-05T07:12:13.939331+00:00"},{"alias_kind":"pith_short_12","alias_value":"SWB3SLGEOZBK","created_at":"2026-07-05T07:12:13.939331+00:00"},{"alias_kind":"pith_short_16","alias_value":"SWB3SLGEOZBK2N4A","created_at":"2026-07-05T07:12:13.939331+00:00"},{"alias_kind":"pith_short_8","alias_value":"SWB3SLGE","created_at":"2026-07-05T07:12:13.939331+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":17,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.29705","citing_title":"GUICrafter: Weakly-Supervised GUI Agent Leveraging Massive Unannotated Screenshots","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2406.16173","citing_title":"Crepe: A Mobile Screen Data Collector Using Graph Query","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2501.16150","citing_title":"A Comprehensive Survey of Agents for Computer Use: Foundations, Challenges, and Future Directions","ref_index":175,"is_internal_anchor":false},{"citing_arxiv_id":"2505.03364","citing_title":"DroidRetriever: A Transparent and Steerable Automation System for Collaborative Mobile Information Seeking","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"2505.16120","citing_title":"LLM-Powered AI Agent Systems and Their Applications in Industry","ref_index":111,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27859","citing_title":"Rethinking Agentic Reinforcement Learning In Large Language Models","ref_index":108,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15542","citing_title":"DRS-GUI: Dynamic Region Search for Training-Free GUI Grounding","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2411.18279","citing_title":"Large Language Model-Brained GUI Agents: A Survey","ref_index":160,"is_internal_anchor":false},{"citing_arxiv_id":"2312.13771","citing_title":"AppAgent: Multimodal Agents as Smartphone Users","ref_index":84,"is_internal_anchor":false},{"citing_arxiv_id":"2401.10935","citing_title":"SeeClick: Harnessing GUI Grounding for Advanced Visual GUI Agents","ref_index":102,"is_internal_anchor":false},{"citing_arxiv_id":"2401.05459","citing_title":"Personal LLM Agents: Insights and Survey about the Capability, Efficiency and Security","ref_index":106,"is_internal_anchor":false},{"citing_arxiv_id":"2512.10371","citing_title":"AgentProg: Empowering Long-Horizon GUI Agents with Program-Guided Context Management","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2412.10345","citing_title":"TraceVLA: Visual Trace Prompting Enhances Spatial-Temporal Awareness for Generalist Robotic Policies","ref_index":99,"is_internal_anchor":false},{"citing_arxiv_id":"2404.07972","citing_title":"OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments","ref_index":56,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23941","citing_title":"GoClick: Lightweight Element Grounding Model for Autonomous GUI Interaction","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27859","citing_title":"Rethinking Agentic Reinforcement Learning In Large Language Models","ref_index":108,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27859","citing_title":"Rethinking Agentic Reinforcement Learning In Large Language Models","ref_index":108,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SWB3SLGEOZBK2N4A3B2BKA6H5Z","json":"https://pith.science/pith/SWB3SLGEOZBK2N4A3B2BKA6H5Z.json","graph_json":"https://pith.science/api/pith-number/SWB3SLGEOZBK2N4A3B2BKA6H5Z/graph.json","events_json":"https://pith.science/api/pith-number/SWB3SLGEOZBK2N4A3B2BKA6H5Z/events.json","paper":"https://pith.science/paper/SWB3SLGE"},"agent_actions":{"view_html":"https://pith.science/pith/SWB3SLGEOZBK2N4A3B2BKA6H5Z","download_json":"https://pith.science/pith/SWB3SLGEOZBK2N4A3B2BKA6H5Z.json","view_paper":"https://pith.science/paper/SWB3SLGE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2311.07562&json=true","fetch_graph":"https://pith.science/api/pith-number/SWB3SLGEOZBK2N4A3B2BKA6H5Z/graph.json","fetch_events":"https://pith.science/api/pith-number/SWB3SLGEOZBK2N4A3B2BKA6H5Z/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SWB3SLGEOZBK2N4A3B2BKA6H5Z/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SWB3SLGEOZBK2N4A3B2BKA6H5Z/action/storage_attestation","attest_author":"https://pith.science/pith/SWB3SLGEOZBK2N4A3B2BKA6H5Z/action/author_attestation","sign_citation":"https://pith.science/pith/SWB3SLGEOZBK2N4A3B2BKA6H5Z/action/citation_signature","submit_replication":"https://pith.science/pith/SWB3SLGEOZBK2N4A3B2BKA6H5Z/action/replication_record"}},"created_at":"2026-07-05T07:12:13.939331+00:00","updated_at":"2026-07-05T07:12:13.939331+00:00"}