{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:5S7QIIHZWBMSOHEHDFBBSVTZYD","short_pith_number":"pith:5S7QIIHZ","canonical_record":{"source":{"id":"2409.07703","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-09-12T02:08:00Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"24764dfb73eb2275270ab1e671fff6f266b0caded3c1d95a53fe9ee68027106f","abstract_canon_sha256":"65a9b710d43d2dc0fd71cc3c580facf85019445ae76d997ee21dd69a3af21630"},"schema_version":"1.0"},"canonical_sha256":"ecbf0420f9b059271c871942195679c0c94e875afb4da22c6534cf2ca5856aa9","source":{"kind":"arxiv","id":"2409.07703","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.07703","created_at":"2026-07-05T10:47:30Z"},{"alias_kind":"arxiv_version","alias_value":"2409.07703v3","created_at":"2026-07-05T10:47:30Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.07703","created_at":"2026-07-05T10:47:30Z"},{"alias_kind":"pith_short_12","alias_value":"5S7QIIHZWBMS","created_at":"2026-07-05T10:47:30Z"},{"alias_kind":"pith_short_16","alias_value":"5S7QIIHZWBMSOHEH","created_at":"2026-07-05T10:47:30Z"},{"alias_kind":"pith_short_8","alias_value":"5S7QIIHZ","created_at":"2026-07-05T10:47:30Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:5S7QIIHZWBMSOHEHDFBBSVTZYD","target":"record","payload":{"canonical_record":{"source":{"id":"2409.07703","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-09-12T02:08:00Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"24764dfb73eb2275270ab1e671fff6f266b0caded3c1d95a53fe9ee68027106f","abstract_canon_sha256":"65a9b710d43d2dc0fd71cc3c580facf85019445ae76d997ee21dd69a3af21630"},"schema_version":"1.0"},"canonical_sha256":"ecbf0420f9b059271c871942195679c0c94e875afb4da22c6534cf2ca5856aa9","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:47:30.800541Z","signature_b64":"ol8IBgEK0teigQavwgsqTpGOlebJlPs9h1VPVARLiiXsTNntdKT+tDF+ochBK3VJpg91ynInHiM1iPLbVvvNDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ecbf0420f9b059271c871942195679c0c94e875afb4da22c6534cf2ca5856aa9","last_reissued_at":"2026-07-05T10:47:30.800062Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:47:30.800062Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2409.07703","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:47:30Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"BlK0LSg77enI7RwCCkgKhh7DZfdPEdtFVoc9g1TfZvsnI1Vip9Y3VxUQJvF7c1DMwkUioa6ZLGAxda6zdnKEAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T05:45:46.593090Z"},"content_sha256":"4759f28ab4701d2151350d17e2a743d25bc5fd154b4a204a94eea9eeb02d3aa9","schema_version":"1.0","event_id":"sha256:4759f28ab4701d2151350d17e2a743d25bc5fd154b4a204a94eea9eeb02d3aa9"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:5S7QIIHZWBMSOHEHDFBBSVTZYD","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"DSBench: How Far Are Data Science Agents from Becoming Data Science Experts?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Dong Yu, Hongming Zhang, Kaixin Ma, Liqiang Jing, Wenhao Yu, Wenlin Yao, Xiaoyang Wang, Xinya Du, Zhehui Huang","submitted_at":"2024-09-12T02:08:00Z","abstract_excerpt":"Large Language Models (LLMs) and Large Vision-Language Models (LVLMs) have demonstrated impressive language/vision reasoning abilities, igniting the recent trend of building agents for targeted applications such as shopping assistants or AI software engineers. Recently, many data science benchmarks have been proposed to investigate their performance in the data science domain. However, existing data science benchmarks still fall short when compared to real-world data science applications due to their simplified settings. To bridge this gap, we introduce DSBench, a comprehensive benchmark desig"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.07703","kind":"arxiv","version":3},"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/2409.07703/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":"2026-07-05T10:47:30Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KEpNNxW2rAF2Cx1hOgKOq3NG4Vyiwg12fm/ZwWqzNXX73vC/Ro9PQKB5QykKjXyoYLf8rulsxXApixYDIkLbBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T05:45:46.593601Z"},"content_sha256":"397ba1f7686018ee0e7ac53be2cbc8c75713b10d0fbb79c487f80d45740b09fe","schema_version":"1.0","event_id":"sha256:397ba1f7686018ee0e7ac53be2cbc8c75713b10d0fbb79c487f80d45740b09fe"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/5S7QIIHZWBMSOHEHDFBBSVTZYD/bundle.json","state_url":"https://pith.science/pith/5S7QIIHZWBMSOHEHDFBBSVTZYD/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/5S7QIIHZWBMSOHEHDFBBSVTZYD/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":"2026-08-07T05:45:46Z","links":{"resolver":"https://pith.science/pith/5S7QIIHZWBMSOHEHDFBBSVTZYD","bundle":"https://pith.science/pith/5S7QIIHZWBMSOHEHDFBBSVTZYD/bundle.json","state":"https://pith.science/pith/5S7QIIHZWBMSOHEHDFBBSVTZYD/state.json","well_known_bundle":"https://pith.science/.well-known/pith/5S7QIIHZWBMSOHEHDFBBSVTZYD/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:5S7QIIHZWBMSOHEHDFBBSVTZYD","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":"65a9b710d43d2dc0fd71cc3c580facf85019445ae76d997ee21dd69a3af21630","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-09-12T02:08:00Z","title_canon_sha256":"24764dfb73eb2275270ab1e671fff6f266b0caded3c1d95a53fe9ee68027106f"},"schema_version":"1.0","source":{"id":"2409.07703","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.07703","created_at":"2026-07-05T10:47:30Z"},{"alias_kind":"arxiv_version","alias_value":"2409.07703v3","created_at":"2026-07-05T10:47:30Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.07703","created_at":"2026-07-05T10:47:30Z"},{"alias_kind":"pith_short_12","alias_value":"5S7QIIHZWBMS","created_at":"2026-07-05T10:47:30Z"},{"alias_kind":"pith_short_16","alias_value":"5S7QIIHZWBMSOHEH","created_at":"2026-07-05T10:47:30Z"},{"alias_kind":"pith_short_8","alias_value":"5S7QIIHZ","created_at":"2026-07-05T10:47:30Z"}],"graph_snapshots":[{"event_id":"sha256:397ba1f7686018ee0e7ac53be2cbc8c75713b10d0fbb79c487f80d45740b09fe","target":"graph","created_at":"2026-07-05T10:47:30Z","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/2409.07703/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) and Large Vision-Language Models (LVLMs) have demonstrated impressive language/vision reasoning abilities, igniting the recent trend of building agents for targeted applications such as shopping assistants or AI software engineers. Recently, many data science benchmarks have been proposed to investigate their performance in the data science domain. However, existing data science benchmarks still fall short when compared to real-world data science applications due to their simplified settings. To bridge this gap, we introduce DSBench, a comprehensive benchmark desig","authors_text":"Dong Yu, Hongming Zhang, Kaixin Ma, Liqiang Jing, Wenhao Yu, Wenlin Yao, Xiaoyang Wang, Xinya Du, Zhehui Huang","cross_cats":["cs.CL"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-09-12T02:08:00Z","title":"DSBench: How Far Are Data Science Agents from Becoming Data Science Experts?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.07703","kind":"arxiv","version":3},"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":"sha256:4759f28ab4701d2151350d17e2a743d25bc5fd154b4a204a94eea9eeb02d3aa9","target":"record","created_at":"2026-07-05T10:47:30Z","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":"65a9b710d43d2dc0fd71cc3c580facf85019445ae76d997ee21dd69a3af21630","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-09-12T02:08:00Z","title_canon_sha256":"24764dfb73eb2275270ab1e671fff6f266b0caded3c1d95a53fe9ee68027106f"},"schema_version":"1.0","source":{"id":"2409.07703","kind":"arxiv","version":3}},"canonical_sha256":"ecbf0420f9b059271c871942195679c0c94e875afb4da22c6534cf2ca5856aa9","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ecbf0420f9b059271c871942195679c0c94e875afb4da22c6534cf2ca5856aa9","first_computed_at":"2026-07-05T10:47:30.800062Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:47:30.800062Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ol8IBgEK0teigQavwgsqTpGOlebJlPs9h1VPVARLiiXsTNntdKT+tDF+ochBK3VJpg91ynInHiM1iPLbVvvNDg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:47:30.800541Z","signed_message":"canonical_sha256_bytes"},"source_id":"2409.07703","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4759f28ab4701d2151350d17e2a743d25bc5fd154b4a204a94eea9eeb02d3aa9","sha256:397ba1f7686018ee0e7ac53be2cbc8c75713b10d0fbb79c487f80d45740b09fe"],"state_sha256":"6de0bff7dfc33892f92a520db6d6767c9d11b544f683ab5ad6bd6fe91a8c6b1a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2c8OwHy18LgmItTh9wMCZM3n7NuNe3x/XWNdVCqNL2Ex4j1wezEc3yK3tSowTIHg/Xya6BxV7Mpuy5b7M9AIBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T05:45:46.599493Z","bundle_sha256":"add141c6fa3fdd32d9196b6722bfd5fd3a6266c4b7a4be2d50ea48accef1d5aa"}}