{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:H56A4BZIF2DD5OMSS437L5Q6TV","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":"d11d7d5d61a3a78aadded7afa2b8cf9f27f1fd5473be8736302f8e33ffea34aa","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-06T17:59:15Z","title_canon_sha256":"9386d3627a0af3eedd2a189bc301e998e282d9a6fc051fef05771093b44f7e13"},"schema_version":"1.0","source":{"id":"2505.03733","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.03733","created_at":"2026-07-05T11:51:26Z"},{"alias_kind":"arxiv_version","alias_value":"2505.03733v2","created_at":"2026-07-05T11:51:26Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.03733","created_at":"2026-07-05T11:51:26Z"},{"alias_kind":"pith_short_12","alias_value":"H56A4BZIF2DD","created_at":"2026-07-05T11:51:26Z"},{"alias_kind":"pith_short_16","alias_value":"H56A4BZIF2DD5OMS","created_at":"2026-07-05T11:51:26Z"},{"alias_kind":"pith_short_8","alias_value":"H56A4BZI","created_at":"2026-07-05T11:51:26Z"}],"graph_snapshots":[{"event_id":"sha256:7cf59908b096ce0e6c0315d7eb79f1b427f2453b815f5b22daa5a883c7da592c","target":"graph","created_at":"2026-07-05T11:51:26Z","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/2505.03733/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"LLM-based agents have demonstrated great potential in generating and managing code within complex codebases. In this paper, we introduce WebGen-Bench, a novel benchmark designed to measure an LLM-based agent's ability to create multi-file website codebases from scratch. It contains diverse instructions for website generation, created through the combined efforts of human annotators and GPT-4o. These instructions span three major categories and thirteen minor categories, encompassing nearly all important types of web applications. To assess the quality of the generated websites, we use GPT-4o t","authors_text":"Aojun Zhou, Han Xiao, Haotian Hou, Hongsheng Li, Houxing Ren, Ke Wang, Mingjie Zhan, Weikang Shi, Yunqiao Yang, Zimu Lu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-06T17:59:15Z","title":"WebGen-Bench: Evaluating LLMs on Generating Interactive and Functional Websites from Scratch"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.03733","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:f9acae5898d9484c966d6dd45899f459a50f38fb4f05c4bc1d26b8ad93f890be","target":"record","created_at":"2026-07-05T11:51:26Z","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":"d11d7d5d61a3a78aadded7afa2b8cf9f27f1fd5473be8736302f8e33ffea34aa","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-06T17:59:15Z","title_canon_sha256":"9386d3627a0af3eedd2a189bc301e998e282d9a6fc051fef05771093b44f7e13"},"schema_version":"1.0","source":{"id":"2505.03733","kind":"arxiv","version":2}},"canonical_sha256":"3f7c0e07282e863eb9929737f5f61e9d4955708efc77e58f5e35263f990a7b68","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3f7c0e07282e863eb9929737f5f61e9d4955708efc77e58f5e35263f990a7b68","first_computed_at":"2026-07-05T11:51:26.682234Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:51:26.682234Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"tnWNaA7RJ/g8f5mLatHSWuqyWOCMuP1Lm6A6NzvtK4v5NJluECCrm1jV9NN3CGZ90FimUj3R+5/Bu6DgHOohCA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:51:26.682739Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.03733","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f9acae5898d9484c966d6dd45899f459a50f38fb4f05c4bc1d26b8ad93f890be","sha256:7cf59908b096ce0e6c0315d7eb79f1b427f2453b815f5b22daa5a883c7da592c"],"state_sha256":"4104719e924a48d8b6edcace05b5f8cc927ae79770247a9a755cfbbb58a011d6"}