{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:EY6T3NRCWOWRMGOJQ5ND5IFVYT","short_pith_number":"pith:EY6T3NRC","schema_version":"1.0","canonical_sha256":"263d3db622b3ad1619c9875a3ea0b5c4f516bcf54d127c3739f052f39cc23b2b","source":{"kind":"arxiv","id":"2507.00938","version":2},"attestation_state":"computed","paper":{"title":"WebArXiv: Evaluating Multimodal Agents on Time-Invariant arXiv Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.DB"],"primary_cat":"cs.IR","authors_text":"Ling Chen, Zihao Sun","submitted_at":"2025-07-01T16:43:57Z","abstract_excerpt":"Recent progress in large language models (LLMs) has enabled the development of autonomous web agents capable of navigating and interacting with real websites. However, evaluating such agents remains challenging due to the instability and inconsistency of existing benchmarks, which often rely on dynamic content or oversimplified simulations. In this work, we introduce WebArXiv, a static and time-invariant benchmark comprising 275 web-based tasks grounded in the arXiv platform. WebArXiv ensures reproducible and reliable evaluation by anchoring tasks in fixed web snapshots with deterministic grou"},"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":"2507.00938","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2025-07-01T16:43:57Z","cross_cats_sorted":["cs.AI","cs.DB"],"title_canon_sha256":"7995e7f61b10a442249b30746762d25a311db6876e2f15be276a4f8698f8b988","abstract_canon_sha256":"4b8903bc3d2e0d590e061653463f4de0f914b0bad0712445a5828dcec01a0a69"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:53:04.944754Z","signature_b64":"7JhcBKWdFDUBwEzY3I07VTQyfxpBQbO+zpr65BxMgDEyr83DKNBlkxduJJNZpNU2jjcaP4fa9SKD7r9QzPfABA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"263d3db622b3ad1619c9875a3ea0b5c4f516bcf54d127c3739f052f39cc23b2b","last_reissued_at":"2026-07-05T11:53:04.944297Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:53:04.944297Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"WebArXiv: Evaluating Multimodal Agents on Time-Invariant arXiv Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.DB"],"primary_cat":"cs.IR","authors_text":"Ling Chen, Zihao Sun","submitted_at":"2025-07-01T16:43:57Z","abstract_excerpt":"Recent progress in large language models (LLMs) has enabled the development of autonomous web agents capable of navigating and interacting with real websites. However, evaluating such agents remains challenging due to the instability and inconsistency of existing benchmarks, which often rely on dynamic content or oversimplified simulations. In this work, we introduce WebArXiv, a static and time-invariant benchmark comprising 275 web-based tasks grounded in the arXiv platform. WebArXiv ensures reproducible and reliable evaluation by anchoring tasks in fixed web snapshots with deterministic grou"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.00938","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/2507.00938/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":"2507.00938","created_at":"2026-07-05T11:53:04.944364+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.00938v2","created_at":"2026-07-05T11:53:04.944364+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.00938","created_at":"2026-07-05T11:53:04.944364+00:00"},{"alias_kind":"pith_short_12","alias_value":"EY6T3NRCWOWR","created_at":"2026-07-05T11:53:04.944364+00:00"},{"alias_kind":"pith_short_16","alias_value":"EY6T3NRCWOWRMGOJ","created_at":"2026-07-05T11:53:04.944364+00:00"},{"alias_kind":"pith_short_8","alias_value":"EY6T3NRC","created_at":"2026-07-05T11:53:04.944364+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/EY6T3NRCWOWRMGOJQ5ND5IFVYT","json":"https://pith.science/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT.json","graph_json":"https://pith.science/api/pith-number/EY6T3NRCWOWRMGOJQ5ND5IFVYT/graph.json","events_json":"https://pith.science/api/pith-number/EY6T3NRCWOWRMGOJQ5ND5IFVYT/events.json","paper":"https://pith.science/paper/EY6T3NRC"},"agent_actions":{"view_html":"https://pith.science/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT","download_json":"https://pith.science/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT.json","view_paper":"https://pith.science/paper/EY6T3NRC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.00938&json=true","fetch_graph":"https://pith.science/api/pith-number/EY6T3NRCWOWRMGOJQ5ND5IFVYT/graph.json","fetch_events":"https://pith.science/api/pith-number/EY6T3NRCWOWRMGOJQ5ND5IFVYT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT/action/storage_attestation","attest_author":"https://pith.science/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT/action/author_attestation","sign_citation":"https://pith.science/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT/action/citation_signature","submit_replication":"https://pith.science/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT/action/replication_record"}},"created_at":"2026-07-05T11:53:04.944364+00:00","updated_at":"2026-07-05T11:53:04.944364+00:00"}