{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:NC5L2GVSEBXHELPPQIZXVC57UB","short_pith_number":"pith:NC5L2GVS","canonical_record":{"source":{"id":"2606.22977","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-22T07:57:53Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f964ef57a424822647422c7a6b51a354a054e8e9c545c6eba5cdb61b6f286fbc","abstract_canon_sha256":"764a78c43f8647079ca62e32a00e58ca3efa5672ac2145149a3a7248c524941f"},"schema_version":"1.0"},"canonical_sha256":"68babd1ab2206e722def82337a8bbfa05d487a2af14050ea19929602c69ac54c","source":{"kind":"arxiv","id":"2606.22977","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.22977","created_at":"2026-06-23T03:14:05Z"},{"alias_kind":"arxiv_version","alias_value":"2606.22977v1","created_at":"2026-06-23T03:14:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.22977","created_at":"2026-06-23T03:14:05Z"},{"alias_kind":"pith_short_12","alias_value":"NC5L2GVSEBXH","created_at":"2026-06-23T03:14:05Z"},{"alias_kind":"pith_short_16","alias_value":"NC5L2GVSEBXHELPP","created_at":"2026-06-23T03:14:05Z"},{"alias_kind":"pith_short_8","alias_value":"NC5L2GVS","created_at":"2026-06-23T03:14:05Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:NC5L2GVSEBXHELPPQIZXVC57UB","target":"record","payload":{"canonical_record":{"source":{"id":"2606.22977","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-22T07:57:53Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f964ef57a424822647422c7a6b51a354a054e8e9c545c6eba5cdb61b6f286fbc","abstract_canon_sha256":"764a78c43f8647079ca62e32a00e58ca3efa5672ac2145149a3a7248c524941f"},"schema_version":"1.0"},"canonical_sha256":"68babd1ab2206e722def82337a8bbfa05d487a2af14050ea19929602c69ac54c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T03:14:05.755348Z","signature_b64":"FuTIomrubKoWy3ff6yFyFhC5mudZzBRFLnDSnYBZQ6z4Q1naZGqHAvaeBUn4Xx5MJOe3hbt5BvHiqjg6lvSPDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"68babd1ab2206e722def82337a8bbfa05d487a2af14050ea19929602c69ac54c","last_reissued_at":"2026-06-23T03:14:05.754962Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T03:14:05.754962Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.22977","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":"2026-06-23T03:14:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6Xb09nchEFytwtvoFiDgEKXJiMidqQLHFKyLEfjALxeQO96OvUySsrQ8WCyRybnUcWatDimmR9aai8ZXCj46DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-04T02:51:55.121820Z"},"content_sha256":"b4b2aedc0965eaca272c60c166dcf92bf9586fbb45391f1c259533cab3e6475f","schema_version":"1.0","event_id":"sha256:b4b2aedc0965eaca272c60c166dcf92bf9586fbb45391f1c259533cab3e6475f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:NC5L2GVSEBXHELPPQIZXVC57UB","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"StatABench: Dataset and Framework for Evaluating Statistical Analysis Capabilities of LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Bingyi Jing, Guanhua Chen, Longyue Wang, Peng Lai, Yixuan Ding, Youxin Zhu","submitted_at":"2026-06-22T07:57:53Z","abstract_excerpt":"Statistical analysis is a broad, complex field requiring both domain knowledge and tool proficiency. While prior work has evaluated large language models (LLMs) in this domain, existing benchmarks remain limited in scope and format. To bridge this gap, we introduce StatABench (Statistical AnalysisBenchmark), a benchmark designed to systematically assess LLMs' statistical analysis capabilities. StatABench comprises two complementary components: Stat-Closed, containing 404 questions across 18 statistical topics in multiple formats (multiple-choice, fill-in-the-blank, decision-making, and practic"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.22977","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/2606.22977/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":"2026-06-23T03:14:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"889OtRsrIxKllOtdQ/nvg3lnyLHhQ1UgWDrI0XZTkht2PoqiVnBwzgru9l9KeFbkoHzZpQ8xM+rIvOX0s2TxCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-04T02:51:55.122193Z"},"content_sha256":"19e92bd566b4a967855704e1d5f15252bd24ee1ee3666eecdeb34f61708c5c7f","schema_version":"1.0","event_id":"sha256:19e92bd566b4a967855704e1d5f15252bd24ee1ee3666eecdeb34f61708c5c7f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/NC5L2GVSEBXHELPPQIZXVC57UB/bundle.json","state_url":"https://pith.science/pith/NC5L2GVSEBXHELPPQIZXVC57UB/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/NC5L2GVSEBXHELPPQIZXVC57UB/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":"2026-07-04T02:51:55Z","links":{"resolver":"https://pith.science/pith/NC5L2GVSEBXHELPPQIZXVC57UB","bundle":"https://pith.science/pith/NC5L2GVSEBXHELPPQIZXVC57UB/bundle.json","state":"https://pith.science/pith/NC5L2GVSEBXHELPPQIZXVC57UB/state.json","well_known_bundle":"https://pith.science/.well-known/pith/NC5L2GVSEBXHELPPQIZXVC57UB/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:NC5L2GVSEBXHELPPQIZXVC57UB","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":"764a78c43f8647079ca62e32a00e58ca3efa5672ac2145149a3a7248c524941f","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-22T07:57:53Z","title_canon_sha256":"f964ef57a424822647422c7a6b51a354a054e8e9c545c6eba5cdb61b6f286fbc"},"schema_version":"1.0","source":{"id":"2606.22977","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.22977","created_at":"2026-06-23T03:14:05Z"},{"alias_kind":"arxiv_version","alias_value":"2606.22977v1","created_at":"2026-06-23T03:14:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.22977","created_at":"2026-06-23T03:14:05Z"},{"alias_kind":"pith_short_12","alias_value":"NC5L2GVSEBXH","created_at":"2026-06-23T03:14:05Z"},{"alias_kind":"pith_short_16","alias_value":"NC5L2GVSEBXHELPP","created_at":"2026-06-23T03:14:05Z"},{"alias_kind":"pith_short_8","alias_value":"NC5L2GVS","created_at":"2026-06-23T03:14:05Z"}],"graph_snapshots":[{"event_id":"sha256:19e92bd566b4a967855704e1d5f15252bd24ee1ee3666eecdeb34f61708c5c7f","target":"graph","created_at":"2026-06-23T03:14:05Z","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/2606.22977/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Statistical analysis is a broad, complex field requiring both domain knowledge and tool proficiency. While prior work has evaluated large language models (LLMs) in this domain, existing benchmarks remain limited in scope and format. To bridge this gap, we introduce StatABench (Statistical AnalysisBenchmark), a benchmark designed to systematically assess LLMs' statistical analysis capabilities. StatABench comprises two complementary components: Stat-Closed, containing 404 questions across 18 statistical topics in multiple formats (multiple-choice, fill-in-the-blank, decision-making, and practic","authors_text":"Bingyi Jing, Guanhua Chen, Longyue Wang, Peng Lai, Yixuan Ding, Youxin Zhu","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-22T07:57:53Z","title":"StatABench: Dataset and Framework for Evaluating Statistical Analysis Capabilities of LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.22977","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:b4b2aedc0965eaca272c60c166dcf92bf9586fbb45391f1c259533cab3e6475f","target":"record","created_at":"2026-06-23T03:14:05Z","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":"764a78c43f8647079ca62e32a00e58ca3efa5672ac2145149a3a7248c524941f","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-22T07:57:53Z","title_canon_sha256":"f964ef57a424822647422c7a6b51a354a054e8e9c545c6eba5cdb61b6f286fbc"},"schema_version":"1.0","source":{"id":"2606.22977","kind":"arxiv","version":1}},"canonical_sha256":"68babd1ab2206e722def82337a8bbfa05d487a2af14050ea19929602c69ac54c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"68babd1ab2206e722def82337a8bbfa05d487a2af14050ea19929602c69ac54c","first_computed_at":"2026-06-23T03:14:05.754962Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-23T03:14:05.754962Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"FuTIomrubKoWy3ff6yFyFhC5mudZzBRFLnDSnYBZQ6z4Q1naZGqHAvaeBUn4Xx5MJOe3hbt5BvHiqjg6lvSPDg==","signature_status":"signed_v1","signed_at":"2026-06-23T03:14:05.755348Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.22977","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b4b2aedc0965eaca272c60c166dcf92bf9586fbb45391f1c259533cab3e6475f","sha256:19e92bd566b4a967855704e1d5f15252bd24ee1ee3666eecdeb34f61708c5c7f"],"state_sha256":"9677232b25ff10bd38d33aa6d58152f0b563cdc42bf3f4b80425cc97c5573413"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4OTTKl0khPBR9htzu9JoNHbFylJyo9tA13o/FxFdX5aztegtOf98a5Uxr20EbNgP1ebXn9cMTJ4KCps8O4dICQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-04T02:51:55.124219Z","bundle_sha256":"d737998959afe777acc9cc3fece673c84a62d7659c3b04824fddc672d419848c"}}