{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:ZVGSGVCIN6NYXA2OWOZXJYJHDX","short_pith_number":"pith:ZVGSGVCI","canonical_record":{"source":{"id":"2406.05761","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-09T12:30:30Z","cross_cats_sorted":[],"title_canon_sha256":"5bf2f5cf80d4abcdae595c65046311823276fc1a65313d7ba49bf3e960970b24","abstract_canon_sha256":"4906a09d6725b79b21d7307dfaa06b830229cac8b8af8b99b1722e101598437f"},"schema_version":"1.0"},"canonical_sha256":"cd4d2354486f9b8b834eb3b374e1271dd404a8c98ad6a6e70d4cfc5adc364686","source":{"kind":"arxiv","id":"2406.05761","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.05761","created_at":"2026-07-05T10:38:33Z"},{"alias_kind":"arxiv_version","alias_value":"2406.05761v2","created_at":"2026-07-05T10:38:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.05761","created_at":"2026-07-05T10:38:33Z"},{"alias_kind":"pith_short_12","alias_value":"ZVGSGVCIN6NY","created_at":"2026-07-05T10:38:33Z"},{"alias_kind":"pith_short_16","alias_value":"ZVGSGVCIN6NYXA2O","created_at":"2026-07-05T10:38:33Z"},{"alias_kind":"pith_short_8","alias_value":"ZVGSGVCI","created_at":"2026-07-05T10:38:33Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:ZVGSGVCIN6NYXA2OWOZXJYJHDX","target":"record","payload":{"canonical_record":{"source":{"id":"2406.05761","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-09T12:30:30Z","cross_cats_sorted":[],"title_canon_sha256":"5bf2f5cf80d4abcdae595c65046311823276fc1a65313d7ba49bf3e960970b24","abstract_canon_sha256":"4906a09d6725b79b21d7307dfaa06b830229cac8b8af8b99b1722e101598437f"},"schema_version":"1.0"},"canonical_sha256":"cd4d2354486f9b8b834eb3b374e1271dd404a8c98ad6a6e70d4cfc5adc364686","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:38:33.996349Z","signature_b64":"aJN9vI/ruVO9S9FtDN5z0zCr0sGjmbD6McZCaxBmHIHyMTQoo+l3c31HNggLT9Qw/wn5hNDshFffjVT8f/D/Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cd4d2354486f9b8b834eb3b374e1271dd404a8c98ad6a6e70d4cfc5adc364686","last_reissued_at":"2026-07-05T10:38:33.995845Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:38:33.995845Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.05761","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":"2026-07-05T10:38:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tb+fUtp9Z0rrMZGtaMa5ruoGJfkaOA6Nxkin/FKYJm7akvr35id8m2mTGNGFqTW0jJR//oxsj8GFPrgn/2O6Cg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-25T21:37:46.120650Z"},"content_sha256":"daa690089563fa37ffd2fae8b2ef0558b95f9aa90cd5ebda0182b724661aa83e","schema_version":"1.0","event_id":"sha256:daa690089563fa37ffd2fae8b2ef0558b95f9aa90cd5ebda0182b724661aa83e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:ZVGSGVCIN6NYXA2OWOZXJYJHDX","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"The BiGGen Bench: A Principled Benchmark for Fine-grained Evaluation of Language Models with Language Models","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bill Yuchen Lin, Chaeeun Kim, Dongkeun Yoon, Graham Neubig, Guijin Son, Haebin Shin, Hanseok Oh, Hyeonbin Hwang, Hyowon Cho, Hyungjoo Chae, Jamin Shin, Jinheon Baek, Jinkyung Jo, Ji Yong Cho, Joel Jang, Juyoung Suk, Kyungjae Lee, Minjoon Seo, Miyoung Ko, Moontae Lee, Namgyu Ho, Noah Lee, Sean Welleck, Se June Joo, Seonghyeon Ye, Seongyun Lee, Seungone Kim, Shayne Longpre, Sheikh Shafayat, Sue Hyun Park, Yejin Cho, Yoonjoo Lee","submitted_at":"2024-06-09T12:30:30Z","abstract_excerpt":"As language models (LMs) become capable of handling a wide range of tasks, their evaluation is becoming as challenging as their development. Most generation benchmarks currently assess LMs using abstract evaluation criteria like helpfulness and harmlessness, which often lack the flexibility and granularity of human assessment. Additionally, these benchmarks tend to focus disproportionately on specific capabilities such as instruction following, leading to coverage bias. To overcome these limitations, we introduce the BiGGen Bench, a principled generation benchmark designed to thoroughly evalua"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.05761","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/2406.05761/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":"2026-07-05T10:38:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7Z8ZW6Faal/VJoTpRb2CDz70JTGag/mycr6twMbSbIwbKu5foSTqfrO6ZlgRjokkWykZVhtY3S+3MoV7mCW3DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-25T21:37:46.121040Z"},"content_sha256":"432097b630099b233d423fa57812e7184865b31b798257e7d8c831893b9b7ba7","schema_version":"1.0","event_id":"sha256:432097b630099b233d423fa57812e7184865b31b798257e7d8c831893b9b7ba7"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ZVGSGVCIN6NYXA2OWOZXJYJHDX/bundle.json","state_url":"https://pith.science/pith/ZVGSGVCIN6NYXA2OWOZXJYJHDX/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ZVGSGVCIN6NYXA2OWOZXJYJHDX/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":"2026-07-25T21:37:46Z","links":{"resolver":"https://pith.science/pith/ZVGSGVCIN6NYXA2OWOZXJYJHDX","bundle":"https://pith.science/pith/ZVGSGVCIN6NYXA2OWOZXJYJHDX/bundle.json","state":"https://pith.science/pith/ZVGSGVCIN6NYXA2OWOZXJYJHDX/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ZVGSGVCIN6NYXA2OWOZXJYJHDX/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:ZVGSGVCIN6NYXA2OWOZXJYJHDX","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":"4906a09d6725b79b21d7307dfaa06b830229cac8b8af8b99b1722e101598437f","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-09T12:30:30Z","title_canon_sha256":"5bf2f5cf80d4abcdae595c65046311823276fc1a65313d7ba49bf3e960970b24"},"schema_version":"1.0","source":{"id":"2406.05761","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.05761","created_at":"2026-07-05T10:38:33Z"},{"alias_kind":"arxiv_version","alias_value":"2406.05761v2","created_at":"2026-07-05T10:38:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.05761","created_at":"2026-07-05T10:38:33Z"},{"alias_kind":"pith_short_12","alias_value":"ZVGSGVCIN6NY","created_at":"2026-07-05T10:38:33Z"},{"alias_kind":"pith_short_16","alias_value":"ZVGSGVCIN6NYXA2O","created_at":"2026-07-05T10:38:33Z"},{"alias_kind":"pith_short_8","alias_value":"ZVGSGVCI","created_at":"2026-07-05T10:38:33Z"}],"graph_snapshots":[{"event_id":"sha256:432097b630099b233d423fa57812e7184865b31b798257e7d8c831893b9b7ba7","target":"graph","created_at":"2026-07-05T10:38:33Z","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/2406.05761/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As language models (LMs) become capable of handling a wide range of tasks, their evaluation is becoming as challenging as their development. Most generation benchmarks currently assess LMs using abstract evaluation criteria like helpfulness and harmlessness, which often lack the flexibility and granularity of human assessment. Additionally, these benchmarks tend to focus disproportionately on specific capabilities such as instruction following, leading to coverage bias. To overcome these limitations, we introduce the BiGGen Bench, a principled generation benchmark designed to thoroughly evalua","authors_text":"Bill Yuchen Lin, Chaeeun Kim, Dongkeun Yoon, Graham Neubig, Guijin Son, Haebin Shin, Hanseok Oh, Hyeonbin Hwang, Hyowon Cho, Hyungjoo Chae, Jamin Shin, Jinheon Baek, Jinkyung Jo, Ji Yong Cho, Joel Jang, Juyoung Suk, Kyungjae Lee, Minjoon Seo, Miyoung Ko, Moontae Lee, Namgyu Ho, Noah Lee, Sean Welleck, Se June Joo, Seonghyeon Ye, Seongyun Lee, Seungone Kim, Shayne Longpre, Sheikh Shafayat, Sue Hyun Park, Yejin Cho, Yoonjoo Lee","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-09T12:30:30Z","title":"The BiGGen Bench: A Principled Benchmark for Fine-grained Evaluation of Language Models with Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.05761","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:daa690089563fa37ffd2fae8b2ef0558b95f9aa90cd5ebda0182b724661aa83e","target":"record","created_at":"2026-07-05T10:38:33Z","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":"4906a09d6725b79b21d7307dfaa06b830229cac8b8af8b99b1722e101598437f","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-09T12:30:30Z","title_canon_sha256":"5bf2f5cf80d4abcdae595c65046311823276fc1a65313d7ba49bf3e960970b24"},"schema_version":"1.0","source":{"id":"2406.05761","kind":"arxiv","version":2}},"canonical_sha256":"cd4d2354486f9b8b834eb3b374e1271dd404a8c98ad6a6e70d4cfc5adc364686","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"cd4d2354486f9b8b834eb3b374e1271dd404a8c98ad6a6e70d4cfc5adc364686","first_computed_at":"2026-07-05T10:38:33.995845Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:38:33.995845Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"aJN9vI/ruVO9S9FtDN5z0zCr0sGjmbD6McZCaxBmHIHyMTQoo+l3c31HNggLT9Qw/wn5hNDshFffjVT8f/D/Dw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:38:33.996349Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.05761","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:daa690089563fa37ffd2fae8b2ef0558b95f9aa90cd5ebda0182b724661aa83e","sha256:432097b630099b233d423fa57812e7184865b31b798257e7d8c831893b9b7ba7"],"state_sha256":"26fea231952287d021c0c828ffcbb1048bc563142b7dc4b6fcc6d4d824e107a4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"d3sSOWJQrdAypEeN15pBF5BUi/wT3NpoohgeZOd3NJvfGwytfE5uo4qm1BjC9jf2oq+nc6foyxQex0pVtIJrCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-25T21:37:46.123137Z","bundle_sha256":"07928bd80b3f055c65e5bbd6e09ac44e8b2148123a2c7992d9974522cde46aa8"}}