{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:LGLYFYS3N5MZN4Z65FGSQVXAWY","short_pith_number":"pith:LGLYFYS3","canonical_record":{"source":{"id":"2502.19187","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-26T14:50:50Z","cross_cats_sorted":[],"title_canon_sha256":"dc9f2ff5c541fa297b52f4390572c0ae609ebc652b8c2c52889e86e4286ff14a","abstract_canon_sha256":"4a7ffd350814a22ac3583bdaefc87b3d423ae0a07e3c27de4f5dba176ffc697d"},"schema_version":"1.0"},"canonical_sha256":"599782e25b6f5996f33ee94d2856e0b60ccb1c84392494a443bf74c5f601dce6","source":{"kind":"arxiv","id":"2502.19187","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.19187","created_at":"2026-07-05T10:59:05Z"},{"alias_kind":"arxiv_version","alias_value":"2502.19187v2","created_at":"2026-07-05T10:59:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.19187","created_at":"2026-07-05T10:59:05Z"},{"alias_kind":"pith_short_12","alias_value":"LGLYFYS3N5MZ","created_at":"2026-07-05T10:59:05Z"},{"alias_kind":"pith_short_16","alias_value":"LGLYFYS3N5MZN4Z6","created_at":"2026-07-05T10:59:05Z"},{"alias_kind":"pith_short_8","alias_value":"LGLYFYS3","created_at":"2026-07-05T10:59:05Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:LGLYFYS3N5MZN4Z65FGSQVXAWY","target":"record","payload":{"canonical_record":{"source":{"id":"2502.19187","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-26T14:50:50Z","cross_cats_sorted":[],"title_canon_sha256":"dc9f2ff5c541fa297b52f4390572c0ae609ebc652b8c2c52889e86e4286ff14a","abstract_canon_sha256":"4a7ffd350814a22ac3583bdaefc87b3d423ae0a07e3c27de4f5dba176ffc697d"},"schema_version":"1.0"},"canonical_sha256":"599782e25b6f5996f33ee94d2856e0b60ccb1c84392494a443bf74c5f601dce6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:59:05.976383Z","signature_b64":"acheoGrJIExWnFkNwTMmVi2/PGLXNeRHsw4jL+NkyA3CpOB071KLDfi+3nwpukHGBCacz0RTKinV6lJZQpYfBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"599782e25b6f5996f33ee94d2856e0b60ccb1c84392494a443bf74c5f601dce6","last_reissued_at":"2026-07-05T10:59:05.975888Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:59:05.975888Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.19187","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":"2026-07-05T10:59:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"p1+eUn09qRkoOkHdiPSRuHe+vbv2LPHBBRc4kNKEjbkKRK79Avo0pNH5+kb/SEDm+KQmmhPN2T93RJMic49RDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T04:14:41.035138Z"},"content_sha256":"1e8cdd07171be83fb946135df09e999b0b8cb5e0f186bfa29459917bc25f5a78","schema_version":"1.0","event_id":"sha256:1e8cdd07171be83fb946135df09e999b0b8cb5e0f186bfa29459917bc25f5a78"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:LGLYFYS3N5MZN4Z65FGSQVXAWY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"BIG-Bench Extra Hard","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bahare Fatemi, Chrysovalantis Anastasiou, Disha Jindal, Gladys Tyen, Hritik Bansal, John Palowitch, Kate Olszewska, Lalit K. Jain, Mehran Kazemi, Nishanth Dikkala, Orhan Firat, Peter Chen, Quoc V. Le, Sanket Vaibhav Mehta, Silvia Chiappa, Uri Shalit, Vinh Q. Tran, Virginia Aglietti, Xin Liu, Yi Tay","submitted_at":"2025-02-26T14:50:50Z","abstract_excerpt":"Large language models (LLMs) are increasingly deployed in everyday applications, demanding robust general reasoning capabilities and diverse reasoning skillset. However, current LLM reasoning benchmarks predominantly focus on mathematical and coding abilities, leaving a gap in evaluating broader reasoning proficiencies. One particular exception is the BIG-Bench dataset, which has served as a crucial benchmark for evaluating the general reasoning capabilities of LLMs, thanks to its diverse set of challenging tasks that allowed for a comprehensive assessment of general reasoning across various s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.19187","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/2502.19187/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":"2026-07-05T10:59:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XzluQxuB+YQld98VwZ7G/cyaAfHM4zc4LarwwRCeeZTipXQYwDo/CU7y8ll0QkVUiIyrXxWM07T8831yLiOlCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T04:14:41.035535Z"},"content_sha256":"816dba013e1355c9c6b7f4e0db4fe00d1bbb59174fc20333398ef6d48d53de87","schema_version":"1.0","event_id":"sha256:816dba013e1355c9c6b7f4e0db4fe00d1bbb59174fc20333398ef6d48d53de87"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/LGLYFYS3N5MZN4Z65FGSQVXAWY/bundle.json","state_url":"https://pith.science/pith/LGLYFYS3N5MZN4Z65FGSQVXAWY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/LGLYFYS3N5MZN4Z65FGSQVXAWY/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":"2026-07-07T04:14:41Z","links":{"resolver":"https://pith.science/pith/LGLYFYS3N5MZN4Z65FGSQVXAWY","bundle":"https://pith.science/pith/LGLYFYS3N5MZN4Z65FGSQVXAWY/bundle.json","state":"https://pith.science/pith/LGLYFYS3N5MZN4Z65FGSQVXAWY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/LGLYFYS3N5MZN4Z65FGSQVXAWY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:LGLYFYS3N5MZN4Z65FGSQVXAWY","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":"4a7ffd350814a22ac3583bdaefc87b3d423ae0a07e3c27de4f5dba176ffc697d","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-26T14:50:50Z","title_canon_sha256":"dc9f2ff5c541fa297b52f4390572c0ae609ebc652b8c2c52889e86e4286ff14a"},"schema_version":"1.0","source":{"id":"2502.19187","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.19187","created_at":"2026-07-05T10:59:05Z"},{"alias_kind":"arxiv_version","alias_value":"2502.19187v2","created_at":"2026-07-05T10:59:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.19187","created_at":"2026-07-05T10:59:05Z"},{"alias_kind":"pith_short_12","alias_value":"LGLYFYS3N5MZ","created_at":"2026-07-05T10:59:05Z"},{"alias_kind":"pith_short_16","alias_value":"LGLYFYS3N5MZN4Z6","created_at":"2026-07-05T10:59:05Z"},{"alias_kind":"pith_short_8","alias_value":"LGLYFYS3","created_at":"2026-07-05T10:59:05Z"}],"graph_snapshots":[{"event_id":"sha256:816dba013e1355c9c6b7f4e0db4fe00d1bbb59174fc20333398ef6d48d53de87","target":"graph","created_at":"2026-07-05T10:59:05Z","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/2502.19187/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) are increasingly deployed in everyday applications, demanding robust general reasoning capabilities and diverse reasoning skillset. However, current LLM reasoning benchmarks predominantly focus on mathematical and coding abilities, leaving a gap in evaluating broader reasoning proficiencies. One particular exception is the BIG-Bench dataset, which has served as a crucial benchmark for evaluating the general reasoning capabilities of LLMs, thanks to its diverse set of challenging tasks that allowed for a comprehensive assessment of general reasoning across various s","authors_text":"Bahare Fatemi, Chrysovalantis Anastasiou, Disha Jindal, Gladys Tyen, Hritik Bansal, John Palowitch, Kate Olszewska, Lalit K. Jain, Mehran Kazemi, Nishanth Dikkala, Orhan Firat, Peter Chen, Quoc V. Le, Sanket Vaibhav Mehta, Silvia Chiappa, Uri Shalit, Vinh Q. Tran, Virginia Aglietti, Xin Liu, Yi Tay","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-26T14:50:50Z","title":"BIG-Bench Extra Hard"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.19187","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":"sha256:1e8cdd07171be83fb946135df09e999b0b8cb5e0f186bfa29459917bc25f5a78","target":"record","created_at":"2026-07-05T10:59:05Z","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":"4a7ffd350814a22ac3583bdaefc87b3d423ae0a07e3c27de4f5dba176ffc697d","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-26T14:50:50Z","title_canon_sha256":"dc9f2ff5c541fa297b52f4390572c0ae609ebc652b8c2c52889e86e4286ff14a"},"schema_version":"1.0","source":{"id":"2502.19187","kind":"arxiv","version":2}},"canonical_sha256":"599782e25b6f5996f33ee94d2856e0b60ccb1c84392494a443bf74c5f601dce6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"599782e25b6f5996f33ee94d2856e0b60ccb1c84392494a443bf74c5f601dce6","first_computed_at":"2026-07-05T10:59:05.975888Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:59:05.975888Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"acheoGrJIExWnFkNwTMmVi2/PGLXNeRHsw4jL+NkyA3CpOB071KLDfi+3nwpukHGBCacz0RTKinV6lJZQpYfBA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:59:05.976383Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.19187","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1e8cdd07171be83fb946135df09e999b0b8cb5e0f186bfa29459917bc25f5a78","sha256:816dba013e1355c9c6b7f4e0db4fe00d1bbb59174fc20333398ef6d48d53de87"],"state_sha256":"ce2222fbfe5750fd3114b5110038b80ce40f140fd61e2acff44880342dd864e5"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"pcdQbZUdCBw+BXeVRdvv2a6wUD1ZCdSgpppU0K6e331x7K98XdnbBEFWDvh30YoQMOTT+lwmKq5S0roex/S4Cw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T04:14:41.037514Z","bundle_sha256":"c7c30a782301dc01a3f86382fd51e0414f7994a2f009c5ac6c2ab3ba7e191b77"}}