{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:VRP2V7XBPDSO3P7CRAV72B77JN","short_pith_number":"pith:VRP2V7XB","schema_version":"1.0","canonical_sha256":"ac5faafee178e4edbfe2882bfd07ff4b564693f2dbdee742ce60c3dc7dd395c9","source":{"kind":"arxiv","id":"2607.22708","version":1},"attestation_state":"computed","paper":{"title":"StepX-Edge: An On-Device UI Vision-Language Model via Architecture-Training-Deployment Co-Design","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Fanyi Wang, Haotian Hu, Jineng Han, Liujian Tang, Wentao Qiu, Yin Wang, Zhenhua Ge","submitted_at":"2026-07-20T14:14:44Z","abstract_excerpt":"Deploying a vision-language model with full UI understanding on end devices has long been trapped between accuracy and efficiency: on one side is the accuracy bar for OCR, screen understanding, visual question answering, and element grounding; on the other is the strict compute, memory, and power budget of mobile chips. Existing work either trades one for the other, or stops at simulation without real-device validation. We present StepX-Edge, a 0.9B-parameter on-device UI vision-language model that resolves this tension through three-layer co-design of architecture, training, and deployment. A"},"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":"2607.22708","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-07-20T14:14:44Z","cross_cats_sorted":[],"title_canon_sha256":"78e8bba7a94a5c70bb9a2673727516c3edbb08e5bf68bc475b4b7e43a505b78a","abstract_canon_sha256":"de01409763cd7a7a51403426225467a968bac4a5952e3013f9a6c3b97be4122a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-28T00:21:50.270170Z","signature_b64":"dUhAw9gjjgwnJFi3u9FPyEvkAMkTjvrfvZG6C6P/XoY4yGCt2Yndc2yFHxkO3OBQnEvvGHE92OKysRES6B5zDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ac5faafee178e4edbfe2882bfd07ff4b564693f2dbdee742ce60c3dc7dd395c9","last_reissued_at":"2026-07-28T00:21:50.269285Z","signature_status":"signed_v1","first_computed_at":"2026-07-28T00:21:50.269285Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"StepX-Edge: An On-Device UI Vision-Language Model via Architecture-Training-Deployment Co-Design","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Fanyi Wang, Haotian Hu, Jineng Han, Liujian Tang, Wentao Qiu, Yin Wang, Zhenhua Ge","submitted_at":"2026-07-20T14:14:44Z","abstract_excerpt":"Deploying a vision-language model with full UI understanding on end devices has long been trapped between accuracy and efficiency: on one side is the accuracy bar for OCR, screen understanding, visual question answering, and element grounding; on the other is the strict compute, memory, and power budget of mobile chips. Existing work either trades one for the other, or stops at simulation without real-device validation. We present StepX-Edge, a 0.9B-parameter on-device UI vision-language model that resolves this tension through three-layer co-design of architecture, training, and deployment. A"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.22708","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/2607.22708/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":"2607.22708","created_at":"2026-07-28T00:21:50.269744+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.22708v1","created_at":"2026-07-28T00:21:50.269744+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.22708","created_at":"2026-07-28T00:21:50.269744+00:00"},{"alias_kind":"pith_short_12","alias_value":"VRP2V7XBPDSO","created_at":"2026-07-28T00:21:50.269744+00:00"},{"alias_kind":"pith_short_16","alias_value":"VRP2V7XBPDSO3P7C","created_at":"2026-07-28T00:21:50.269744+00:00"},{"alias_kind":"pith_short_8","alias_value":"VRP2V7XB","created_at":"2026-07-28T00:21:50.269744+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/VRP2V7XBPDSO3P7CRAV72B77JN","json":"https://pith.science/pith/VRP2V7XBPDSO3P7CRAV72B77JN.json","graph_json":"https://pith.science/api/pith-number/VRP2V7XBPDSO3P7CRAV72B77JN/graph.json","events_json":"https://pith.science/api/pith-number/VRP2V7XBPDSO3P7CRAV72B77JN/events.json","paper":"https://pith.science/paper/VRP2V7XB"},"agent_actions":{"view_html":"https://pith.science/pith/VRP2V7XBPDSO3P7CRAV72B77JN","download_json":"https://pith.science/pith/VRP2V7XBPDSO3P7CRAV72B77JN.json","view_paper":"https://pith.science/paper/VRP2V7XB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.22708&json=true","fetch_graph":"https://pith.science/api/pith-number/VRP2V7XBPDSO3P7CRAV72B77JN/graph.json","fetch_events":"https://pith.science/api/pith-number/VRP2V7XBPDSO3P7CRAV72B77JN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VRP2V7XBPDSO3P7CRAV72B77JN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VRP2V7XBPDSO3P7CRAV72B77JN/action/storage_attestation","attest_author":"https://pith.science/pith/VRP2V7XBPDSO3P7CRAV72B77JN/action/author_attestation","sign_citation":"https://pith.science/pith/VRP2V7XBPDSO3P7CRAV72B77JN/action/citation_signature","submit_replication":"https://pith.science/pith/VRP2V7XBPDSO3P7CRAV72B77JN/action/replication_record"}},"created_at":"2026-07-28T00:21:50.269744+00:00","updated_at":"2026-07-28T00:21:50.269744+00:00"}