{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:WJ2RKY2JURHL3H3ZVLO2FVOQE5","short_pith_number":"pith:WJ2RKY2J","canonical_record":{"source":{"id":"2412.17933","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-23T19:45:20Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"802769c0b04beb8f530788d8ead14f16628eb528eff69323780fabf6c4701b98","abstract_canon_sha256":"adfbe719ada11530f20c4776be52d470c798e872efdd262f66131524d8f07738"},"schema_version":"1.0"},"canonical_sha256":"b275156349a44ebd9f79aadda2d5d0276b5a790d25d9bc84f15ed3f2c377faee","source":{"kind":"arxiv","id":"2412.17933","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.17933","created_at":"2026-07-05T11:07:18Z"},{"alias_kind":"arxiv_version","alias_value":"2412.17933v2","created_at":"2026-07-05T11:07:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.17933","created_at":"2026-07-05T11:07:18Z"},{"alias_kind":"pith_short_12","alias_value":"WJ2RKY2JURHL","created_at":"2026-07-05T11:07:18Z"},{"alias_kind":"pith_short_16","alias_value":"WJ2RKY2JURHL3H3Z","created_at":"2026-07-05T11:07:18Z"},{"alias_kind":"pith_short_8","alias_value":"WJ2RKY2J","created_at":"2026-07-05T11:07:18Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:WJ2RKY2JURHL3H3ZVLO2FVOQE5","target":"record","payload":{"canonical_record":{"source":{"id":"2412.17933","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-23T19:45:20Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"802769c0b04beb8f530788d8ead14f16628eb528eff69323780fabf6c4701b98","abstract_canon_sha256":"adfbe719ada11530f20c4776be52d470c798e872efdd262f66131524d8f07738"},"schema_version":"1.0"},"canonical_sha256":"b275156349a44ebd9f79aadda2d5d0276b5a790d25d9bc84f15ed3f2c377faee","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:07:18.482167Z","signature_b64":"QYVB37WjmRxTri4jH1XuVZZx0hdWrBGYXmvzP8A7/t18680q+6LJKmhuYU5xy4nGZuNvSMNvoo1+LWyLnCogBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b275156349a44ebd9f79aadda2d5d0276b5a790d25d9bc84f15ed3f2c377faee","last_reissued_at":"2026-07-05T11:07:18.481656Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:07:18.481656Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.17933","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":"2026-07-05T11:07:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YP2IB30LtFByRCBL3pZg2XO6cPM15dVuzPhLY+MvFalfjfjtoPVLrkfTYyKmcidy2aH/Op38UyZtmIJbzGVzDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-12T22:37:48.416363Z"},"content_sha256":"6281911de3baf2871b2de81204c002608ea503f4690ebf576794a794bcad0331","schema_version":"1.0","event_id":"sha256:6281911de3baf2871b2de81204c002608ea503f4690ebf576794a794bcad0331"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:WJ2RKY2JURHL3H3ZVLO2FVOQE5","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"BenCzechMark : A Czech-centric Multitask and Multimetric Benchmark for Large Language Models with Duel Scoring Mechanism","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Adam Jirkovsky, Ales Horak, Alexander Polok, David Adamczyk, Hynek Kydlicek, Jan Dolezal, Jan Hula, Jan Kapsa, Karel Bene\\v{s}, Karel Ondrej, Martin Docekal, Martin Fajcik, Michal Hradis, Michal Stefanik, Pavel Smrz, Petr Hyner, Radoslav Sabol, Zuzana Neverilova","submitted_at":"2024-12-23T19:45:20Z","abstract_excerpt":"We present BenCzechMark (BCM), the first comprehensive Czech language benchmark designed for large language models, offering diverse tasks, multiple task formats, and multiple evaluation metrics. Its duel scoring system is grounded in statistical significance theory and uses aggregation across tasks inspired by social preference theory. Our benchmark encompasses 50 challenging tasks, with corresponding test datasets, primarily in native Czech, with 14 newly collected ones. These tasks span 8 categories and cover diverse domains, including historical Czech news, essays from pupils or language l"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.17933","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/2412.17933/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":"2026-07-05T11:07:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MLlo+vX/C1HJhV9/jH0uOHL6dU01U8wB+4vvXzVWSMJc0l0QypE9aQLW7jKGEFKciPPPoEsC5Ffvt8WROVytAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-12T22:37:48.416781Z"},"content_sha256":"856e9ff019b0343f50d5c4cd4b1f19789353f2c170268ab6cc79d49605b43ad5","schema_version":"1.0","event_id":"sha256:856e9ff019b0343f50d5c4cd4b1f19789353f2c170268ab6cc79d49605b43ad5"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/WJ2RKY2JURHL3H3ZVLO2FVOQE5/bundle.json","state_url":"https://pith.science/pith/WJ2RKY2JURHL3H3ZVLO2FVOQE5/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/WJ2RKY2JURHL3H3ZVLO2FVOQE5/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":"2026-08-12T22:37:48Z","links":{"resolver":"https://pith.science/pith/WJ2RKY2JURHL3H3ZVLO2FVOQE5","bundle":"https://pith.science/pith/WJ2RKY2JURHL3H3ZVLO2FVOQE5/bundle.json","state":"https://pith.science/pith/WJ2RKY2JURHL3H3ZVLO2FVOQE5/state.json","well_known_bundle":"https://pith.science/.well-known/pith/WJ2RKY2JURHL3H3ZVLO2FVOQE5/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:WJ2RKY2JURHL3H3ZVLO2FVOQE5","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":"adfbe719ada11530f20c4776be52d470c798e872efdd262f66131524d8f07738","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-23T19:45:20Z","title_canon_sha256":"802769c0b04beb8f530788d8ead14f16628eb528eff69323780fabf6c4701b98"},"schema_version":"1.0","source":{"id":"2412.17933","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.17933","created_at":"2026-07-05T11:07:18Z"},{"alias_kind":"arxiv_version","alias_value":"2412.17933v2","created_at":"2026-07-05T11:07:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.17933","created_at":"2026-07-05T11:07:18Z"},{"alias_kind":"pith_short_12","alias_value":"WJ2RKY2JURHL","created_at":"2026-07-05T11:07:18Z"},{"alias_kind":"pith_short_16","alias_value":"WJ2RKY2JURHL3H3Z","created_at":"2026-07-05T11:07:18Z"},{"alias_kind":"pith_short_8","alias_value":"WJ2RKY2J","created_at":"2026-07-05T11:07:18Z"}],"graph_snapshots":[{"event_id":"sha256:856e9ff019b0343f50d5c4cd4b1f19789353f2c170268ab6cc79d49605b43ad5","target":"graph","created_at":"2026-07-05T11:07:18Z","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/2412.17933/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We present BenCzechMark (BCM), the first comprehensive Czech language benchmark designed for large language models, offering diverse tasks, multiple task formats, and multiple evaluation metrics. Its duel scoring system is grounded in statistical significance theory and uses aggregation across tasks inspired by social preference theory. Our benchmark encompasses 50 challenging tasks, with corresponding test datasets, primarily in native Czech, with 14 newly collected ones. These tasks span 8 categories and cover diverse domains, including historical Czech news, essays from pupils or language l","authors_text":"Adam Jirkovsky, Ales Horak, Alexander Polok, David Adamczyk, Hynek Kydlicek, Jan Dolezal, Jan Hula, Jan Kapsa, Karel Bene\\v{s}, Karel Ondrej, Martin Docekal, Martin Fajcik, Michal Hradis, Michal Stefanik, Pavel Smrz, Petr Hyner, Radoslav Sabol, Zuzana Neverilova","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-23T19:45:20Z","title":"BenCzechMark : A Czech-centric Multitask and Multimetric Benchmark for Large Language Models with Duel Scoring Mechanism"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.17933","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":"sha256:6281911de3baf2871b2de81204c002608ea503f4690ebf576794a794bcad0331","target":"record","created_at":"2026-07-05T11:07:18Z","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":"adfbe719ada11530f20c4776be52d470c798e872efdd262f66131524d8f07738","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-23T19:45:20Z","title_canon_sha256":"802769c0b04beb8f530788d8ead14f16628eb528eff69323780fabf6c4701b98"},"schema_version":"1.0","source":{"id":"2412.17933","kind":"arxiv","version":2}},"canonical_sha256":"b275156349a44ebd9f79aadda2d5d0276b5a790d25d9bc84f15ed3f2c377faee","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b275156349a44ebd9f79aadda2d5d0276b5a790d25d9bc84f15ed3f2c377faee","first_computed_at":"2026-07-05T11:07:18.481656Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:07:18.481656Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"QYVB37WjmRxTri4jH1XuVZZx0hdWrBGYXmvzP8A7/t18680q+6LJKmhuYU5xy4nGZuNvSMNvoo1+LWyLnCogBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:07:18.482167Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.17933","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6281911de3baf2871b2de81204c002608ea503f4690ebf576794a794bcad0331","sha256:856e9ff019b0343f50d5c4cd4b1f19789353f2c170268ab6cc79d49605b43ad5"],"state_sha256":"c1a4bfe430c1346e18ea897fe523654ef1d43a34af8492ed6e39436bfb76076f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PLjeI26q2TPA75oEbXfiG+XP537yrjbWDhMOtEp24lUx26C96LQuJFg7JupD7QKkzJWG6UrzNeTiw5GC0KOYDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-12T22:37:48.419479Z","bundle_sha256":"f3f6ae9c29c19441959dab6a1ad50dd40f180c0f6001381d10b9fc53394f828b"}}