{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:FJUAZ43DVWATRS2KL4DZ2U4JBX","short_pith_number":"pith:FJUAZ43D","schema_version":"1.0","canonical_sha256":"2a680cf363ad8138cb4a5f079d53890dc3e2f81ff939495d4f5448e64fd39d24","source":{"kind":"arxiv","id":"2607.17050","version":1},"attestation_state":"computed","paper":{"title":"EvoGUI: An Evolution-Aware Benchmark for GUI State-Transition Understanding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Borui Zhang, Jie Zhou, Jiwen Lu, Minglei Shi, Yaohan Yang","submitted_at":"2026-07-19T03:29:08Z","abstract_excerpt":"GUI agents must reason about how actions transform interface states, but end-to-end success rates entangle this ability with perception, grounding, planning, and recovery. We introduce EvoGUI, a diagnostic framework that converts normalized GUI trajectories into three complementary visual question answering probes: temporal ordering, inverse action/value prediction, and contrastive one-step successor discrimination. Their labels are derived from trajectory order and logged actions, requiring no additional task-label annotation after trajectory normalization. We instantiate EvoGUI-Bench from Mi"},"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":"2607.17050","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-07-19T03:29:08Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"eebeb0c8b11fc3b229a40bc061a88d1381a31052755567ae0fd2b063e2267c76","abstract_canon_sha256":"75cada30219a92a992e409e4c0ee5256dd0dfa03987a4e73b2d1d5004d048cc9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-21T01:21:12.244457Z","signature_b64":"sB1gPFYkOQsRh8pW/cVYTQKg4vdIXen3gg4RYs46jh+jNcKYiv/kLgWU+fUQC3MgbZUWP37ztASWrvs4bfk0BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2a680cf363ad8138cb4a5f079d53890dc3e2f81ff939495d4f5448e64fd39d24","last_reissued_at":"2026-07-21T01:21:12.243592Z","signature_status":"signed_v1","first_computed_at":"2026-07-21T01:21:12.243592Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"EvoGUI: An Evolution-Aware Benchmark for GUI State-Transition Understanding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Borui Zhang, Jie Zhou, Jiwen Lu, Minglei Shi, Yaohan Yang","submitted_at":"2026-07-19T03:29:08Z","abstract_excerpt":"GUI agents must reason about how actions transform interface states, but end-to-end success rates entangle this ability with perception, grounding, planning, and recovery. We introduce EvoGUI, a diagnostic framework that converts normalized GUI trajectories into three complementary visual question answering probes: temporal ordering, inverse action/value prediction, and contrastive one-step successor discrimination. Their labels are derived from trajectory order and logged actions, requiring no additional task-label annotation after trajectory normalization. We instantiate EvoGUI-Bench from Mi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.17050","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/2607.17050/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":"2607.17050","created_at":"2026-07-21T01:21:12.244039+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.17050v1","created_at":"2026-07-21T01:21:12.244039+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.17050","created_at":"2026-07-21T01:21:12.244039+00:00"},{"alias_kind":"pith_short_12","alias_value":"FJUAZ43DVWAT","created_at":"2026-07-21T01:21:12.244039+00:00"},{"alias_kind":"pith_short_16","alias_value":"FJUAZ43DVWATRS2K","created_at":"2026-07-21T01:21:12.244039+00:00"},{"alias_kind":"pith_short_8","alias_value":"FJUAZ43D","created_at":"2026-07-21T01:21:12.244039+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/FJUAZ43DVWATRS2KL4DZ2U4JBX","json":"https://pith.science/pith/FJUAZ43DVWATRS2KL4DZ2U4JBX.json","graph_json":"https://pith.science/api/pith-number/FJUAZ43DVWATRS2KL4DZ2U4JBX/graph.json","events_json":"https://pith.science/api/pith-number/FJUAZ43DVWATRS2KL4DZ2U4JBX/events.json","paper":"https://pith.science/paper/FJUAZ43D"},"agent_actions":{"view_html":"https://pith.science/pith/FJUAZ43DVWATRS2KL4DZ2U4JBX","download_json":"https://pith.science/pith/FJUAZ43DVWATRS2KL4DZ2U4JBX.json","view_paper":"https://pith.science/paper/FJUAZ43D","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.17050&json=true","fetch_graph":"https://pith.science/api/pith-number/FJUAZ43DVWATRS2KL4DZ2U4JBX/graph.json","fetch_events":"https://pith.science/api/pith-number/FJUAZ43DVWATRS2KL4DZ2U4JBX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FJUAZ43DVWATRS2KL4DZ2U4JBX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FJUAZ43DVWATRS2KL4DZ2U4JBX/action/storage_attestation","attest_author":"https://pith.science/pith/FJUAZ43DVWATRS2KL4DZ2U4JBX/action/author_attestation","sign_citation":"https://pith.science/pith/FJUAZ43DVWATRS2KL4DZ2U4JBX/action/citation_signature","submit_replication":"https://pith.science/pith/FJUAZ43DVWATRS2KL4DZ2U4JBX/action/replication_record"}},"created_at":"2026-07-21T01:21:12.244039+00:00","updated_at":"2026-07-21T01:21:12.244039+00:00"}