{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:TBPEPHBCRT4HRNCN4BTFSHLAZF","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"5eaba51c9902f3be14d8c2206e59e2a774d5df5b559a746a598cebd013899c46","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.HC","submitted_at":"2024-04-12T15:39:09Z","title_canon_sha256":"72801d16d0d985fb93d312297ac0452255fe601554d315e78af3662961f99eb1"},"schema_version":"1.0","source":{"id":"2404.16054","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.16054","created_at":"2026-07-05T08:51:16Z"},{"alias_kind":"arxiv_version","alias_value":"2404.16054v2","created_at":"2026-07-05T08:51:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.16054","created_at":"2026-07-05T08:51:16Z"},{"alias_kind":"pith_short_12","alias_value":"TBPEPHBCRT4H","created_at":"2026-07-05T08:51:16Z"},{"alias_kind":"pith_short_16","alias_value":"TBPEPHBCRT4HRNCN","created_at":"2026-07-05T08:51:16Z"},{"alias_kind":"pith_short_8","alias_value":"TBPEPHBC","created_at":"2026-07-05T08:51:16Z"}],"graph_snapshots":[{"event_id":"sha256:7e8508c3b3a251d85ef677e761d3a4894a8b58660413474d2bce43459d189719","target":"graph","created_at":"2026-07-05T08:51:16Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2404.16054/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The emergent large language/multimodal models facilitate the evolution of mobile agents, especially in mobile UI task automation. However, existing evaluation approaches, which rely on human validation or established datasets to compare agent-predicted actions with predefined action sequences, are unscalable and unfaithful. To overcome these limitations, this paper presents LlamaTouch, a testbed for on-device mobile UI task execution and faithful, scalable task evaluation. By observing that the task execution process only transfers UI states, LlamaTouch employs a novel evaluation approach that","authors_text":"Li Zhang, Longxi Gao, Mengwei Xu, Shihe Wang, Xianqing Jia, Yuanchun Li, Yunhe Yan, Zhihan Zheng","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.HC","submitted_at":"2024-04-12T15:39:09Z","title":"LlamaTouch: A Faithful and Scalable Testbed for Mobile UI Task Automation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.16054","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:84873ed00d31d300023f2fe4be4b2fa6ed7e0a8de6f14ae234bac6a460820d20","target":"record","created_at":"2026-07-05T08:51:16Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"5eaba51c9902f3be14d8c2206e59e2a774d5df5b559a746a598cebd013899c46","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.HC","submitted_at":"2024-04-12T15:39:09Z","title_canon_sha256":"72801d16d0d985fb93d312297ac0452255fe601554d315e78af3662961f99eb1"},"schema_version":"1.0","source":{"id":"2404.16054","kind":"arxiv","version":2}},"canonical_sha256":"985e479c228cf878b44de066591d60c94bb495123cd56d6583946da67464df11","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"985e479c228cf878b44de066591d60c94bb495123cd56d6583946da67464df11","first_computed_at":"2026-07-05T08:51:16.743313Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:51:16.743313Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"VgMMUU1fIS/vSR5zThFHjzA6nRecHQcYd24kfTd9KzyYRuinEExcjE4Zu99psWKuTlaiE2RGKaOOpmHcEVg/Bg==","signature_status":"signed_v1","signed_at":"2026-07-05T08:51:16.743802Z","signed_message":"canonical_sha256_bytes"},"source_id":"2404.16054","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:84873ed00d31d300023f2fe4be4b2fa6ed7e0a8de6f14ae234bac6a460820d20","sha256:7e8508c3b3a251d85ef677e761d3a4894a8b58660413474d2bce43459d189719"],"state_sha256":"ab1fafd733c5354cdcd034e691c855d259052ba7965058e34174dcc62b5b82ce"}