{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:JO5N76H6EQHCZV3RASJZWQCBRP","short_pith_number":"pith:JO5N76H6","canonical_record":{"source":{"id":"2605.30000","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-28T14:30:33Z","cross_cats_sorted":[],"title_canon_sha256":"544b2e30faaad0e943d5e62541602298ab844fefa670b7533677f93216bf4ffb","abstract_canon_sha256":"53b3212211a2eda3cd872070f4663f40906bc475bb3cf821769e25bd0dfff1df"},"schema_version":"1.0"},"canonical_sha256":"4bbadff8fe240e2cd77104939b40418be0dc5c924ab7264e0ad55ab70e71f5df","source":{"kind":"arxiv","id":"2605.30000","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.30000","created_at":"2026-05-29T02:06:06Z"},{"alias_kind":"arxiv_version","alias_value":"2605.30000v1","created_at":"2026-05-29T02:06:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.30000","created_at":"2026-05-29T02:06:06Z"},{"alias_kind":"pith_short_12","alias_value":"JO5N76H6EQHC","created_at":"2026-05-29T02:06:06Z"},{"alias_kind":"pith_short_16","alias_value":"JO5N76H6EQHCZV3R","created_at":"2026-05-29T02:06:06Z"},{"alias_kind":"pith_short_8","alias_value":"JO5N76H6","created_at":"2026-05-29T02:06:06Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:JO5N76H6EQHCZV3RASJZWQCBRP","target":"record","payload":{"canonical_record":{"source":{"id":"2605.30000","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-28T14:30:33Z","cross_cats_sorted":[],"title_canon_sha256":"544b2e30faaad0e943d5e62541602298ab844fefa670b7533677f93216bf4ffb","abstract_canon_sha256":"53b3212211a2eda3cd872070f4663f40906bc475bb3cf821769e25bd0dfff1df"},"schema_version":"1.0"},"canonical_sha256":"4bbadff8fe240e2cd77104939b40418be0dc5c924ab7264e0ad55ab70e71f5df","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T02:06:06.250245Z","signature_b64":"iXCNXJuDbnxqSTtvUELfbUPWkXqA1sXfs4U9i8B2OVI3zbMjVL+ITCXa0+OZKRSBTWYEQpKD6OyCD0N6B1lTAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4bbadff8fe240e2cd77104939b40418be0dc5c924ab7264e0ad55ab70e71f5df","last_reissued_at":"2026-05-29T02:06:06.249311Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T02:06:06.249311Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.30000","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-29T02:06:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7IU2CU3g0UScNwkvj7DaS/2jh5uyoir47gddHmf9Q2e/k1gJUIMaivMVynlmGaNHpxCu6MmccrwPT5Dtyar3AQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T12:08:19.272106Z"},"content_sha256":"31838a9d7f3a3dcf80eee03bf96cc16fa7c10d27836696a41c883c784db4c417","schema_version":"1.0","event_id":"sha256:31838a9d7f3a3dcf80eee03bf96cc16fa7c10d27836696a41c883c784db4c417"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:JO5N76H6EQHCZV3RASJZWQCBRP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Cookie-Bench: Continuous On-screen Key Interaction Evaluation for Web Generation","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Fan Ding, Hangting Lou, Haoqing Yu, Haoyue Yang, Hua Wu, Jing Liu, Jingyao Li, Siqi Bao, Yifeng Kou, Zhangxiao Shen, Zhengfan Wu","submitted_at":"2026-05-28T14:30:33Z","abstract_excerpt":"Front-end web code has become a core product surface for every frontier LLM release, yet evaluating these interactive applications at development speed remains costly because human-judged leaderboards like Arena do not scale. Existing automated proxies typically lean on reference implementations, test suites, or rigid checklists, and tend to miss the reasoned synthesis a human reviewer performs over a live session. We articulate a new evaluation regime that is simultaneously reference-free, autonomously driven, and holistically reasoned, and instantiate it through two artifacts. \\textbf{\\datan"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.30000","kind":"arxiv","version":1},"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/2605.30000/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":"2026-05-29T02:06:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GqBQNrSgMamroHQPicIfC7snri31kqFP1+qbFO8nnMCCO6AVanX5MBLkI/xpLQXv4tp14ES2/QaYbHv4Hb8sCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T12:08:19.272875Z"},"content_sha256":"03da5d4e785ef68a25e6b5150386f5bb12e4ca5598bd8b4de5f4288dd1eac8b0","schema_version":"1.0","event_id":"sha256:03da5d4e785ef68a25e6b5150386f5bb12e4ca5598bd8b4de5f4288dd1eac8b0"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JO5N76H6EQHCZV3RASJZWQCBRP/bundle.json","state_url":"https://pith.science/pith/JO5N76H6EQHCZV3RASJZWQCBRP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JO5N76H6EQHCZV3RASJZWQCBRP/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":"2026-05-30T12:08:19Z","links":{"resolver":"https://pith.science/pith/JO5N76H6EQHCZV3RASJZWQCBRP","bundle":"https://pith.science/pith/JO5N76H6EQHCZV3RASJZWQCBRP/bundle.json","state":"https://pith.science/pith/JO5N76H6EQHCZV3RASJZWQCBRP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JO5N76H6EQHCZV3RASJZWQCBRP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:JO5N76H6EQHCZV3RASJZWQCBRP","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":"53b3212211a2eda3cd872070f4663f40906bc475bb3cf821769e25bd0dfff1df","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-28T14:30:33Z","title_canon_sha256":"544b2e30faaad0e943d5e62541602298ab844fefa670b7533677f93216bf4ffb"},"schema_version":"1.0","source":{"id":"2605.30000","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.30000","created_at":"2026-05-29T02:06:06Z"},{"alias_kind":"arxiv_version","alias_value":"2605.30000v1","created_at":"2026-05-29T02:06:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.30000","created_at":"2026-05-29T02:06:06Z"},{"alias_kind":"pith_short_12","alias_value":"JO5N76H6EQHC","created_at":"2026-05-29T02:06:06Z"},{"alias_kind":"pith_short_16","alias_value":"JO5N76H6EQHCZV3R","created_at":"2026-05-29T02:06:06Z"},{"alias_kind":"pith_short_8","alias_value":"JO5N76H6","created_at":"2026-05-29T02:06:06Z"}],"graph_snapshots":[{"event_id":"sha256:03da5d4e785ef68a25e6b5150386f5bb12e4ca5598bd8b4de5f4288dd1eac8b0","target":"graph","created_at":"2026-05-29T02:06:06Z","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/2605.30000/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Front-end web code has become a core product surface for every frontier LLM release, yet evaluating these interactive applications at development speed remains costly because human-judged leaderboards like Arena do not scale. Existing automated proxies typically lean on reference implementations, test suites, or rigid checklists, and tend to miss the reasoned synthesis a human reviewer performs over a live session. We articulate a new evaluation regime that is simultaneously reference-free, autonomously driven, and holistically reasoned, and instantiate it through two artifacts. \\textbf{\\datan","authors_text":"Fan Ding, Hangting Lou, Haoqing Yu, Haoyue Yang, Hua Wu, Jing Liu, Jingyao Li, Siqi Bao, Yifeng Kou, Zhangxiao Shen, Zhengfan Wu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-28T14:30:33Z","title":"Cookie-Bench: Continuous On-screen Key Interaction Evaluation for Web Generation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.30000","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":"sha256:31838a9d7f3a3dcf80eee03bf96cc16fa7c10d27836696a41c883c784db4c417","target":"record","created_at":"2026-05-29T02:06:06Z","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":"53b3212211a2eda3cd872070f4663f40906bc475bb3cf821769e25bd0dfff1df","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-28T14:30:33Z","title_canon_sha256":"544b2e30faaad0e943d5e62541602298ab844fefa670b7533677f93216bf4ffb"},"schema_version":"1.0","source":{"id":"2605.30000","kind":"arxiv","version":1}},"canonical_sha256":"4bbadff8fe240e2cd77104939b40418be0dc5c924ab7264e0ad55ab70e71f5df","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4bbadff8fe240e2cd77104939b40418be0dc5c924ab7264e0ad55ab70e71f5df","first_computed_at":"2026-05-29T02:06:06.249311Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-29T02:06:06.249311Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"iXCNXJuDbnxqSTtvUELfbUPWkXqA1sXfs4U9i8B2OVI3zbMjVL+ITCXa0+OZKRSBTWYEQpKD6OyCD0N6B1lTAg==","signature_status":"signed_v1","signed_at":"2026-05-29T02:06:06.250245Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.30000","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:31838a9d7f3a3dcf80eee03bf96cc16fa7c10d27836696a41c883c784db4c417","sha256:03da5d4e785ef68a25e6b5150386f5bb12e4ca5598bd8b4de5f4288dd1eac8b0"],"state_sha256":"6e5224f25aa1255fb358fafdbc2b1606b1ec8a5d837d73c75c444e146c36588b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gkpZHPwPlev1w7m/MV0lsRirlZN18m/xouWluPQ9DN+ldw4rUs8hWcIymUqGfpN95R1ZgEvbixheTB1zdtF/Bw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-30T12:08:19.276087Z","bundle_sha256":"b79a948013c0206b8c199122b3fa9c7bf447f3fc9c8a2475f9fd11f4b197bd2f"}}