{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:OSJFKTE6CTZE35Y3IYZBO7SDRI","short_pith_number":"pith:OSJFKTE6","schema_version":"1.0","canonical_sha256":"7492554c9e14f24df71b4632177e438a15512812f31b21b78efc64b328dd4de3","source":{"kind":"arxiv","id":"2607.20536","version":1},"attestation_state":"computed","paper":{"title":"AppWorld-UL: Benchmarking Diverse Agent-User Interactions for Tool-Use","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG","cs.MA"],"primary_cat":"cs.AI","authors_text":"Ashish Sabharwal, Harsh Trivedi, Jane Pan, Junzhi Chen, Michael JQ Zhang, Niranjan Balasubramanian, Tejas Srinivasan","submitted_at":"2026-07-10T16:55:04Z","abstract_excerpt":"Tool-use agents that address day-to-day digital tasks such as ordering groceries must not only operate applications, but also interact with the user, e.g., to ask clarification questions, prompt for confirmation, and inform the user when the instruction is infeasible. However, current benchmarks for evaluating agent-user interactions do not capture the diversity of such interactions. Further, they operate in small environments with few, often non-state-changing, APIs. To address this gap, we introduce AppWorld-UL, a ``user-in-the-loop'' benchmark of 516 challenging tasks requiring diverse agen"},"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":"2607.20536","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-10T16:55:04Z","cross_cats_sorted":["cs.CL","cs.LG","cs.MA"],"title_canon_sha256":"9e222ceb6f96ee22e038ba30e8ca636a151639d0790809b59d6ee4f59aae8028","abstract_canon_sha256":"2c97b2601d8fb92c6678f57e5a4ff6c99f64905e4c6ec53074996bf8a1307a01"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-24T00:23:21.175156Z","signature_b64":"dZLmlL3cB3Fe6k9BwLHbN7Vjil2vmEQ+A89a8RGxtJBDPbJ0edUXlG3qYrxj095FOEUwoS4XNp3SRzEJ0nulBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7492554c9e14f24df71b4632177e438a15512812f31b21b78efc64b328dd4de3","last_reissued_at":"2026-07-24T00:23:21.172665Z","signature_status":"signed_v1","first_computed_at":"2026-07-24T00:23:21.172665Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AppWorld-UL: Benchmarking Diverse Agent-User Interactions for Tool-Use","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG","cs.MA"],"primary_cat":"cs.AI","authors_text":"Ashish Sabharwal, Harsh Trivedi, Jane Pan, Junzhi Chen, Michael JQ Zhang, Niranjan Balasubramanian, Tejas Srinivasan","submitted_at":"2026-07-10T16:55:04Z","abstract_excerpt":"Tool-use agents that address day-to-day digital tasks such as ordering groceries must not only operate applications, but also interact with the user, e.g., to ask clarification questions, prompt for confirmation, and inform the user when the instruction is infeasible. However, current benchmarks for evaluating agent-user interactions do not capture the diversity of such interactions. Further, they operate in small environments with few, often non-state-changing, APIs. To address this gap, we introduce AppWorld-UL, a ``user-in-the-loop'' benchmark of 516 challenging tasks requiring diverse agen"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.20536","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/2607.20536/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":"2607.20536","created_at":"2026-07-24T00:23:21.174571+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.20536v1","created_at":"2026-07-24T00:23:21.174571+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.20536","created_at":"2026-07-24T00:23:21.174571+00:00"},{"alias_kind":"pith_short_12","alias_value":"OSJFKTE6CTZE","created_at":"2026-07-24T00:23:21.174571+00:00"},{"alias_kind":"pith_short_16","alias_value":"OSJFKTE6CTZE35Y3","created_at":"2026-07-24T00:23:21.174571+00:00"},{"alias_kind":"pith_short_8","alias_value":"OSJFKTE6","created_at":"2026-07-24T00:23:21.174571+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/OSJFKTE6CTZE35Y3IYZBO7SDRI","json":"https://pith.science/pith/OSJFKTE6CTZE35Y3IYZBO7SDRI.json","graph_json":"https://pith.science/api/pith-number/OSJFKTE6CTZE35Y3IYZBO7SDRI/graph.json","events_json":"https://pith.science/api/pith-number/OSJFKTE6CTZE35Y3IYZBO7SDRI/events.json","paper":"https://pith.science/paper/OSJFKTE6"},"agent_actions":{"view_html":"https://pith.science/pith/OSJFKTE6CTZE35Y3IYZBO7SDRI","download_json":"https://pith.science/pith/OSJFKTE6CTZE35Y3IYZBO7SDRI.json","view_paper":"https://pith.science/paper/OSJFKTE6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.20536&json=true","fetch_graph":"https://pith.science/api/pith-number/OSJFKTE6CTZE35Y3IYZBO7SDRI/graph.json","fetch_events":"https://pith.science/api/pith-number/OSJFKTE6CTZE35Y3IYZBO7SDRI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OSJFKTE6CTZE35Y3IYZBO7SDRI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OSJFKTE6CTZE35Y3IYZBO7SDRI/action/storage_attestation","attest_author":"https://pith.science/pith/OSJFKTE6CTZE35Y3IYZBO7SDRI/action/author_attestation","sign_citation":"https://pith.science/pith/OSJFKTE6CTZE35Y3IYZBO7SDRI/action/citation_signature","submit_replication":"https://pith.science/pith/OSJFKTE6CTZE35Y3IYZBO7SDRI/action/replication_record"}},"created_at":"2026-07-24T00:23:21.174571+00:00","updated_at":"2026-07-24T00:23:21.174571+00:00"}