{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:G34JKKB3X2Z7PTNI73XD2ZBDAN","short_pith_number":"pith:G34JKKB3","schema_version":"1.0","canonical_sha256":"36f895283bbeb3f7cda8feee3d6423035d63d991c50d0eb9a971eb1ad3af49fe","source":{"kind":"arxiv","id":"2412.19723","version":3},"attestation_state":"computed","paper":{"title":"OS-Genesis: Automating GUI Agent Trajectory Construction via Reverse Task Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.CV","cs.HC"],"primary_cat":"cs.AI","authors_text":"Ben Kao, Chengyou Jia, Chuanyang Jin, Fangzhi Xu, Guohao Li, Junxian He, Kanzhi Cheng, Liheng Chen, Qiushi Sun, Yian Wang, Yu Qiao, Zhenyu Wu, Zhiyong Wu, Zhoumianze Liu, Zichen Ding","submitted_at":"2024-12-27T16:21:58Z","abstract_excerpt":"Graphical User Interface (GUI) agents powered by Vision-Language Models (VLMs) have demonstrated human-like computer control capability. Despite their utility in advancing digital automation, a critical bottleneck persists: collecting high-quality trajectory data for training. Common practices for collecting such data rely on human supervision or synthetic data generation through executing pre-defined tasks, which are either resource-intensive or unable to guarantee data quality. Moreover, these methods suffer from limited data diversity and significant gaps between synthetic data and real-wor"},"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":"2412.19723","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-12-27T16:21:58Z","cross_cats_sorted":["cs.CL","cs.CV","cs.HC"],"title_canon_sha256":"acd0d72f7dcd56c3da1ed0d1bc0c350e43ea03fc6ff5081fb73dd87d99c5c806","abstract_canon_sha256":"27ef6b25ca5578651771b96ad4b39f94b58f4087ce5cc518f75b749b6e885131"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:28:13.267547Z","signature_b64":"7bIVeb5G0xm3ZasWshM68b5AuLRiyPM3SVziUan7wm3XGUmxiK9L6zg6D21kX0M2fzPHVhpIdc5s/4bb0SXBAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"36f895283bbeb3f7cda8feee3d6423035d63d991c50d0eb9a971eb1ad3af49fe","last_reissued_at":"2026-07-05T11:28:13.266998Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:28:13.266998Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"OS-Genesis: Automating GUI Agent Trajectory Construction via Reverse Task Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.CV","cs.HC"],"primary_cat":"cs.AI","authors_text":"Ben Kao, Chengyou Jia, Chuanyang Jin, Fangzhi Xu, Guohao Li, Junxian He, Kanzhi Cheng, Liheng Chen, Qiushi Sun, Yian Wang, Yu Qiao, Zhenyu Wu, Zhiyong Wu, Zhoumianze Liu, Zichen Ding","submitted_at":"2024-12-27T16:21:58Z","abstract_excerpt":"Graphical User Interface (GUI) agents powered by Vision-Language Models (VLMs) have demonstrated human-like computer control capability. Despite their utility in advancing digital automation, a critical bottleneck persists: collecting high-quality trajectory data for training. Common practices for collecting such data rely on human supervision or synthetic data generation through executing pre-defined tasks, which are either resource-intensive or unable to guarantee data quality. Moreover, these methods suffer from limited data diversity and significant gaps between synthetic data and real-wor"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.19723","kind":"arxiv","version":3},"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/2412.19723/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":"2412.19723","created_at":"2026-07-05T11:28:13.267066+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.19723v3","created_at":"2026-07-05T11:28:13.267066+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.19723","created_at":"2026-07-05T11:28:13.267066+00:00"},{"alias_kind":"pith_short_12","alias_value":"G34JKKB3X2Z7","created_at":"2026-07-05T11:28:13.267066+00:00"},{"alias_kind":"pith_short_16","alias_value":"G34JKKB3X2Z7PTNI","created_at":"2026-07-05T11:28:13.267066+00:00"},{"alias_kind":"pith_short_8","alias_value":"G34JKKB3","created_at":"2026-07-05T11:28:13.267066+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":12,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21654","citing_title":"ChainWorld: Composing Long-Horizon Desktop Workloads from Atomic OSWorld Tasks","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11520","citing_title":"ISE: An Execution-Grounded Recipe for Multi-Turn OS-Agent Trajectories","ref_index":57,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09447","citing_title":"AliyunConsoleAgent: Training Web Agents in Real-World Cloud Environments via Distillation and Reinforcement Learning","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31365","citing_title":"Learning to Adapt: Self-Improving Web Agent via Cognitive-Aware Exploration","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2506.09373","citing_title":"LPO: Towards Accurate GUI Agent Interaction via Location Preference Optimization","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2506.20332","citing_title":"Mobile-R1: Towards Interactive Capability for VLM-Based Mobile Agent via Systematic Training","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2509.21982","citing_title":"RISK: A Framework for GUI Agents in E-commerce Risk Management","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2511.06101","citing_title":"SynthAgent: Adapting Web Agents with Synthetic Supervision","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2504.19678","citing_title":"From LLM Reasoning to Autonomous AI Agents: A Comprehensive Review","ref_index":153,"is_internal_anchor":false},{"citing_arxiv_id":"2507.21046","citing_title":"A Survey of Self-Evolving Agents: What, When, How, and Where to Evolve on the Path to Artificial Super Intelligence","ref_index":244,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13822","citing_title":"UI-Copilot: Advancing Long-Horizon GUI Automation via Tool-Integrated Policy Optimization","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14956","citing_title":"FedGUI: Benchmarking Federated GUI Agents across Heterogeneous Platforms, Devices, and Operating Systems","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/G34JKKB3X2Z7PTNI73XD2ZBDAN","json":"https://pith.science/pith/G34JKKB3X2Z7PTNI73XD2ZBDAN.json","graph_json":"https://pith.science/api/pith-number/G34JKKB3X2Z7PTNI73XD2ZBDAN/graph.json","events_json":"https://pith.science/api/pith-number/G34JKKB3X2Z7PTNI73XD2ZBDAN/events.json","paper":"https://pith.science/paper/G34JKKB3"},"agent_actions":{"view_html":"https://pith.science/pith/G34JKKB3X2Z7PTNI73XD2ZBDAN","download_json":"https://pith.science/pith/G34JKKB3X2Z7PTNI73XD2ZBDAN.json","view_paper":"https://pith.science/paper/G34JKKB3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.19723&json=true","fetch_graph":"https://pith.science/api/pith-number/G34JKKB3X2Z7PTNI73XD2ZBDAN/graph.json","fetch_events":"https://pith.science/api/pith-number/G34JKKB3X2Z7PTNI73XD2ZBDAN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/G34JKKB3X2Z7PTNI73XD2ZBDAN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/G34JKKB3X2Z7PTNI73XD2ZBDAN/action/storage_attestation","attest_author":"https://pith.science/pith/G34JKKB3X2Z7PTNI73XD2ZBDAN/action/author_attestation","sign_citation":"https://pith.science/pith/G34JKKB3X2Z7PTNI73XD2ZBDAN/action/citation_signature","submit_replication":"https://pith.science/pith/G34JKKB3X2Z7PTNI73XD2ZBDAN/action/replication_record"}},"created_at":"2026-07-05T11:28:13.267066+00:00","updated_at":"2026-07-05T11:28:13.267066+00:00"}