{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:WUTSVMJINFWJR24R4HXUJHO5ZF","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":"713576a384e64557f8d9365e33662c3cf60d861076bc40cf90bf26e66ad6f527","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-04-09T02:29:39Z","title_canon_sha256":"a413b1d90e7119e3703aaabac932d21672586306d4efe788164a8dcb5604bbd9"},"schema_version":"1.0","source":{"id":"2404.05955","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.05955","created_at":"2026-07-05T08:06:01Z"},{"alias_kind":"arxiv_version","alias_value":"2404.05955v1","created_at":"2026-07-05T08:06:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.05955","created_at":"2026-07-05T08:06:01Z"},{"alias_kind":"pith_short_12","alias_value":"WUTSVMJINFWJ","created_at":"2026-07-05T08:06:01Z"},{"alias_kind":"pith_short_16","alias_value":"WUTSVMJINFWJR24R","created_at":"2026-07-05T08:06:01Z"},{"alias_kind":"pith_short_8","alias_value":"WUTSVMJI","created_at":"2026-07-05T08:06:01Z"}],"graph_snapshots":[{"event_id":"sha256:0c75323303947be0e51adaa68af9d135b7323bfbfe975b9b0adbb6f3a7f18e4e","target":"graph","created_at":"2026-07-05T08:06:01Z","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/2404.05955/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multimodal Large Language models (MLLMs) have shown promise in web-related tasks, but evaluating their performance in the web domain remains a challenge due to the lack of comprehensive benchmarks. Existing benchmarks are either designed for general multimodal tasks, failing to capture the unique characteristics of web pages, or focus on end-to-end web agent tasks, unable to measure fine-grained abilities such as OCR, understanding, and grounding. In this paper, we introduce \\bench{}, a multimodal benchmark designed to assess the capabilities of MLLMs across a variety of web tasks. \\bench{} co","authors_text":"Bill Yuchen Lin, Graham Neubig, Junpeng Liu, Wai Lam, Xiang Yue, Yifan Song, Yuanzhi Li","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-04-09T02:29:39Z","title":"VisualWebBench: How Far Have Multimodal LLMs Evolved in Web Page Understanding and Grounding?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.05955","kind":"arxiv","version":1},"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":"sha256:711d2cba088fac39ff5508cc3755cd2fc771f4966761292ec2961a6fb8fee18a","target":"record","created_at":"2026-07-05T08:06:01Z","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":"713576a384e64557f8d9365e33662c3cf60d861076bc40cf90bf26e66ad6f527","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-04-09T02:29:39Z","title_canon_sha256":"a413b1d90e7119e3703aaabac932d21672586306d4efe788164a8dcb5604bbd9"},"schema_version":"1.0","source":{"id":"2404.05955","kind":"arxiv","version":1}},"canonical_sha256":"b5272ab128696c98eb91e1ef449dddc9561ed867e9465c31b8b435378f3340a8","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b5272ab128696c98eb91e1ef449dddc9561ed867e9465c31b8b435378f3340a8","first_computed_at":"2026-07-05T08:06:01.584806Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:06:01.584806Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"YIPv3D0Ev/DziQJnRvl6Akv5SX61i7EZJ3lMlkh3VJ1DzaEdzmmGG3ipwpVi2YZ9PjP//RF/pmfvAlEXa4rxCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:06:01.585265Z","signed_message":"canonical_sha256_bytes"},"source_id":"2404.05955","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:711d2cba088fac39ff5508cc3755cd2fc771f4966761292ec2961a6fb8fee18a","sha256:0c75323303947be0e51adaa68af9d135b7323bfbfe975b9b0adbb6f3a7f18e4e"],"state_sha256":"004f580739113a20d5b439d7d7702040a8540e4843002f296f0d47cb7849f223"}