{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ENLSPLV2VBZC5Z2CYJXCFHK4YZ","short_pith_number":"pith:ENLSPLV2","schema_version":"1.0","canonical_sha256":"235727aebaa8722ee742c26e229d5cc66e0198f436d5fb81fa73ac8e5a2e6200","source":{"kind":"arxiv","id":"2510.09872","version":2},"attestation_state":"computed","paper":{"title":"WARC-Bench: Web Archive Based Benchmark for GUI Subtask Executions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Charlene Y. Lee, Cheng Chang, Gang Li, Ignacio Cases, Manpreet Kaur, Peng Qi, Rishu Garg, Sanjari Srivastava, Yanan Xie, Yining Mao, Yuezhang Li","submitted_at":"2025-10-10T21:20:51Z","abstract_excerpt":"Training web agents to navigate complex, real-world websites requires them to master $\\textit{subtasks}$ - short-horizon interactions on multiple UI components (e.g., choosing the correct date in a date picker, or scrolling in a container to extract information). We introduce WARC-Bench (Web Archive Benchmark), a novel web navigation benchmark featuring 438 tasks designed to evaluate multimodal AI agents on subtasks. WARC-Bench enables sandboxed interactions with dynamic and realistic webpages using Web ARChive files. We show that WARC-Bench is challenging for leading computer-use models, with"},"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":"2510.09872","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-10T21:20:51Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"bb3790d6f7234d3504fac410e90dc27d7e0fab9e79c48dc77ac1d6f3632db177","abstract_canon_sha256":"56f08b56ac11b34970ff0149b743534d398ca842719913534b1585f4fbeb268f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:04:59.904722Z","signature_b64":"Zy5NiN+e1qVT3ehJQ5pK3VHn5DETdQnXqQd3SiQ7IpVAcgwcDyTWVXka3w15y47STtU/mBbZExl1mGASvZGDAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"235727aebaa8722ee742c26e229d5cc66e0198f436d5fb81fa73ac8e5a2e6200","last_reissued_at":"2026-05-20T01:04:59.903716Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:04:59.903716Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"WARC-Bench: Web Archive Based Benchmark for GUI Subtask Executions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Charlene Y. Lee, Cheng Chang, Gang Li, Ignacio Cases, Manpreet Kaur, Peng Qi, Rishu Garg, Sanjari Srivastava, Yanan Xie, Yining Mao, Yuezhang Li","submitted_at":"2025-10-10T21:20:51Z","abstract_excerpt":"Training web agents to navigate complex, real-world websites requires them to master $\\textit{subtasks}$ - short-horizon interactions on multiple UI components (e.g., choosing the correct date in a date picker, or scrolling in a container to extract information). We introduce WARC-Bench (Web Archive Benchmark), a novel web navigation benchmark featuring 438 tasks designed to evaluate multimodal AI agents on subtasks. WARC-Bench enables sandboxed interactions with dynamic and realistic webpages using Web ARChive files. We show that WARC-Bench is challenging for leading computer-use models, with"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.09872","kind":"arxiv","version":2},"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/2510.09872/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":"2510.09872","created_at":"2026-05-20T01:04:59.903861+00:00"},{"alias_kind":"arxiv_version","alias_value":"2510.09872v2","created_at":"2026-05-20T01:04:59.903861+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.09872","created_at":"2026-05-20T01:04:59.903861+00:00"},{"alias_kind":"pith_short_12","alias_value":"ENLSPLV2VBZC","created_at":"2026-05-20T01:04:59.903861+00:00"},{"alias_kind":"pith_short_16","alias_value":"ENLSPLV2VBZC5Z2C","created_at":"2026-05-20T01:04:59.903861+00:00"},{"alias_kind":"pith_short_8","alias_value":"ENLSPLV2","created_at":"2026-05-20T01:04:59.903861+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/ENLSPLV2VBZC5Z2CYJXCFHK4YZ","json":"https://pith.science/pith/ENLSPLV2VBZC5Z2CYJXCFHK4YZ.json","graph_json":"https://pith.science/api/pith-number/ENLSPLV2VBZC5Z2CYJXCFHK4YZ/graph.json","events_json":"https://pith.science/api/pith-number/ENLSPLV2VBZC5Z2CYJXCFHK4YZ/events.json","paper":"https://pith.science/paper/ENLSPLV2"},"agent_actions":{"view_html":"https://pith.science/pith/ENLSPLV2VBZC5Z2CYJXCFHK4YZ","download_json":"https://pith.science/pith/ENLSPLV2VBZC5Z2CYJXCFHK4YZ.json","view_paper":"https://pith.science/paper/ENLSPLV2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2510.09872&json=true","fetch_graph":"https://pith.science/api/pith-number/ENLSPLV2VBZC5Z2CYJXCFHK4YZ/graph.json","fetch_events":"https://pith.science/api/pith-number/ENLSPLV2VBZC5Z2CYJXCFHK4YZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ENLSPLV2VBZC5Z2CYJXCFHK4YZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ENLSPLV2VBZC5Z2CYJXCFHK4YZ/action/storage_attestation","attest_author":"https://pith.science/pith/ENLSPLV2VBZC5Z2CYJXCFHK4YZ/action/author_attestation","sign_citation":"https://pith.science/pith/ENLSPLV2VBZC5Z2CYJXCFHK4YZ/action/citation_signature","submit_replication":"https://pith.science/pith/ENLSPLV2VBZC5Z2CYJXCFHK4YZ/action/replication_record"}},"created_at":"2026-05-20T01:04:59.903861+00:00","updated_at":"2026-05-20T01:04:59.903861+00:00"}