{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:EY6T3NRCWOWRMGOJQ5ND5IFVYT","short_pith_number":"pith:EY6T3NRC","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"},"canonical_sha256":"263d3db622b3ad1619c9875a3ea0b5c4f516bcf54d127c3739f052f39cc23b2b","source":{"kind":"arxiv","id":"2507.00938","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.00938","created_at":"2026-07-05T11:53:04Z"},{"alias_kind":"arxiv_version","alias_value":"2507.00938v2","created_at":"2026-07-05T11:53:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.00938","created_at":"2026-07-05T11:53:04Z"},{"alias_kind":"pith_short_12","alias_value":"EY6T3NRCWOWR","created_at":"2026-07-05T11:53:04Z"},{"alias_kind":"pith_short_16","alias_value":"EY6T3NRCWOWRMGOJ","created_at":"2026-07-05T11:53:04Z"},{"alias_kind":"pith_short_8","alias_value":"EY6T3NRC","created_at":"2026-07-05T11:53:04Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:EY6T3NRCWOWRMGOJQ5ND5IFVYT","target":"record","payload":{"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"},"canonical_sha256":"263d3db622b3ad1619c9875a3ea0b5c4f516bcf54d127c3739f052f39cc23b2b","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"},"source_kind":"arxiv","source_id":"2507.00938","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:53:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"658/4/TzIl6L90bcKJ7xaMRSrWRXpckT+CW0sfmgq4AqOqwE4tjc9CiS66FyEONzJvAm4tMesLJeTFZBEfehBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T08:40:15.463660Z"},"content_sha256":"80fadb7841db9bf40acafdf3f0372534425361f80ba1388cade44c6dca480b7a","schema_version":"1.0","event_id":"sha256:80fadb7841db9bf40acafdf3f0372534425361f80ba1388cade44c6dca480b7a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:EY6T3NRCWOWRMGOJQ5ND5IFVYT","target":"graph","payload":{"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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:53:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"WQ3MADst6QnR5Jb0k9JQ+VRqbuGGK992SCWa3fcSTevMyiVPFVRuCEJD+Sa+DgvVenal1BbQCMipBQonq6cBCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T08:40:15.464722Z"},"content_sha256":"584f101b26a5b2b1773d71a3e3ee11244b29b164a9630072fbe08a63f595d800","schema_version":"1.0","event_id":"sha256:584f101b26a5b2b1773d71a3e3ee11244b29b164a9630072fbe08a63f595d800"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT/bundle.json","state_url":"https://pith.science/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-10T08:40:15Z","links":{"resolver":"https://pith.science/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT","bundle":"https://pith.science/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT/bundle.json","state":"https://pith.science/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT/state.json","well_known_bundle":"https://pith.science/.well-known/pith/EY6T3NRCWOWRMGOJQ5ND5IFVYT/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:EY6T3NRCWOWRMGOJQ5ND5IFVYT","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":"4b8903bc3d2e0d590e061653463f4de0f914b0bad0712445a5828dcec01a0a69","cross_cats_sorted":["cs.AI","cs.DB"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2025-07-01T16:43:57Z","title_canon_sha256":"7995e7f61b10a442249b30746762d25a311db6876e2f15be276a4f8698f8b988"},"schema_version":"1.0","source":{"id":"2507.00938","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.00938","created_at":"2026-07-05T11:53:04Z"},{"alias_kind":"arxiv_version","alias_value":"2507.00938v2","created_at":"2026-07-05T11:53:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.00938","created_at":"2026-07-05T11:53:04Z"},{"alias_kind":"pith_short_12","alias_value":"EY6T3NRCWOWR","created_at":"2026-07-05T11:53:04Z"},{"alias_kind":"pith_short_16","alias_value":"EY6T3NRCWOWRMGOJ","created_at":"2026-07-05T11:53:04Z"},{"alias_kind":"pith_short_8","alias_value":"EY6T3NRC","created_at":"2026-07-05T11:53:04Z"}],"graph_snapshots":[{"event_id":"sha256:584f101b26a5b2b1773d71a3e3ee11244b29b164a9630072fbe08a63f595d800","target":"graph","created_at":"2026-07-05T11:53:04Z","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/2507.00938/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"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","authors_text":"Ling Chen, Zihao Sun","cross_cats":["cs.AI","cs.DB"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2025-07-01T16:43:57Z","title":"WebArXiv: Evaluating Multimodal Agents on Time-Invariant arXiv Tasks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.00938","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":"sha256:80fadb7841db9bf40acafdf3f0372534425361f80ba1388cade44c6dca480b7a","target":"record","created_at":"2026-07-05T11:53:04Z","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":"4b8903bc3d2e0d590e061653463f4de0f914b0bad0712445a5828dcec01a0a69","cross_cats_sorted":["cs.AI","cs.DB"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2025-07-01T16:43:57Z","title_canon_sha256":"7995e7f61b10a442249b30746762d25a311db6876e2f15be276a4f8698f8b988"},"schema_version":"1.0","source":{"id":"2507.00938","kind":"arxiv","version":2}},"canonical_sha256":"263d3db622b3ad1619c9875a3ea0b5c4f516bcf54d127c3739f052f39cc23b2b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"263d3db622b3ad1619c9875a3ea0b5c4f516bcf54d127c3739f052f39cc23b2b","first_computed_at":"2026-07-05T11:53:04.944297Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:53:04.944297Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"7JhcBKWdFDUBwEzY3I07VTQyfxpBQbO+zpr65BxMgDEyr83DKNBlkxduJJNZpNU2jjcaP4fa9SKD7r9QzPfABA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:53:04.944754Z","signed_message":"canonical_sha256_bytes"},"source_id":"2507.00938","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:80fadb7841db9bf40acafdf3f0372534425361f80ba1388cade44c6dca480b7a","sha256:584f101b26a5b2b1773d71a3e3ee11244b29b164a9630072fbe08a63f595d800"],"state_sha256":"61ccde0cb7f9e18f02c32a75d365fbe36c03e91e8613017b70ca1df0aad56644"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"lKNJz4fYBx6q6RrRJhuUJuB+7nqrZQjg3ApyuSgv5rroyjFUwwTuae4i+t3abmCF89vMK8ZwwZvuBkcZvbZZDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T08:40:15.477314Z","bundle_sha256":"ccc1de399c424d7c121c6759a78675d907512f0b3acf0e274d98374e792c1448"}}