{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:WICDBVJPGSVTFYFV56JZ7MXI6Z","short_pith_number":"pith:WICDBVJP","canonical_record":{"source":{"id":"2401.05507","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-01-10T19:04:00Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"040523bacb11169d07d114f4b82f0261c04cf4a3a45d5b3e0d7f76797b0a9e32","abstract_canon_sha256":"649e56ea8500ab3edfc8aa5ee3f2fe2edcc5365e4f0d8a8a68054d1aacb1611b"},"schema_version":"1.0"},"canonical_sha256":"b20430d52f34ab32e0b5ef939fb2e8f6575307eeaf7cbce6b586ea71aa298a6f","source":{"kind":"arxiv","id":"2401.05507","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2401.05507","created_at":"2026-07-05T07:54:35Z"},{"alias_kind":"arxiv_version","alias_value":"2401.05507v3","created_at":"2026-07-05T07:54:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.05507","created_at":"2026-07-05T07:54:35Z"},{"alias_kind":"pith_short_12","alias_value":"WICDBVJPGSVT","created_at":"2026-07-05T07:54:35Z"},{"alias_kind":"pith_short_16","alias_value":"WICDBVJPGSVTFYFV","created_at":"2026-07-05T07:54:35Z"},{"alias_kind":"pith_short_8","alias_value":"WICDBVJP","created_at":"2026-07-05T07:54:35Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:WICDBVJPGSVTFYFV56JZ7MXI6Z","target":"record","payload":{"canonical_record":{"source":{"id":"2401.05507","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-01-10T19:04:00Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"040523bacb11169d07d114f4b82f0261c04cf4a3a45d5b3e0d7f76797b0a9e32","abstract_canon_sha256":"649e56ea8500ab3edfc8aa5ee3f2fe2edcc5365e4f0d8a8a68054d1aacb1611b"},"schema_version":"1.0"},"canonical_sha256":"b20430d52f34ab32e0b5ef939fb2e8f6575307eeaf7cbce6b586ea71aa298a6f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:54:35.318276Z","signature_b64":"8k0wj6sOoB02oTQmWFMBQc8bJe9KST6zG51SwqDMtFiVvYMc9vsB570wY4oYkrnkQNLEyF77kc1HklKCnloCCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b20430d52f34ab32e0b5ef939fb2e8f6575307eeaf7cbce6b586ea71aa298a6f","last_reissued_at":"2026-07-05T07:54:35.317781Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:54:35.317781Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2401.05507","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":"2026-07-05T07:54:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"n5hfiDRhY7dHHCrxUqFYe5Pb544XgjvuE5nqi1jubbcKV6hfFB6piqI+jbFzvEL2zDXAFxNhb3FNqZAw5uPVBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T13:49:04.406919Z"},"content_sha256":"06435f973513f2e2c2d3af9493bd5fd1b0e5f99c20d88b7b5ead11e0211472b8","schema_version":"1.0","event_id":"sha256:06435f973513f2e2c2d3af9493bd5fd1b0e5f99c20d88b7b5ead11e0211472b8"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:WICDBVJPGSVTFYFV56JZ7MXI6Z","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"InfiAgent-DABench: Evaluating Agents on Data Analysis Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Fei Wu, Guoyin Wang, Hongxia Yang, Jianbo Yuan, Jingjing Xu, Jing Su, Jiwei Li, Kun Kuang, Ming Zhu, Qianli Ma, Shuang Wei, Xueyu Hu, Xuwu Wang, Yang Yang, Yao Cheng, Ziwei Chai, Ziyu Zhao","submitted_at":"2024-01-10T19:04:00Z","abstract_excerpt":"In this paper, we introduce InfiAgent-DABench, the first benchmark specifically designed to evaluate LLM-based agents on data analysis tasks. These tasks require agents to end-to-end solving complex tasks by interacting with an execution environment. This benchmark contains DAEval, a dataset consisting of 257 data analysis questions derived from 52 CSV files, and an agent framework which incorporates LLMs to serve as data analysis agents for both serving and evaluation. Since data analysis questions are often open-ended and hard to evaluate without human supervision, we adopt a format-promptin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.05507","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/2401.05507/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":"2026-07-05T07:54:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2FMqY7LYyWWWsZ7QElL0eglgtrjaj2dg0wM+ZpFopusn2ZOi6/g7kige90r6dTkyfzOsSeQ2gIJHOs5Q9d1kBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T13:49:04.407483Z"},"content_sha256":"863e8745a11881b96a2c511be3361dc73d484bb21c785f2e2172f28e99d07df1","schema_version":"1.0","event_id":"sha256:863e8745a11881b96a2c511be3361dc73d484bb21c785f2e2172f28e99d07df1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/WICDBVJPGSVTFYFV56JZ7MXI6Z/bundle.json","state_url":"https://pith.science/pith/WICDBVJPGSVTFYFV56JZ7MXI6Z/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/WICDBVJPGSVTFYFV56JZ7MXI6Z/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":"2026-08-04T13:49:04Z","links":{"resolver":"https://pith.science/pith/WICDBVJPGSVTFYFV56JZ7MXI6Z","bundle":"https://pith.science/pith/WICDBVJPGSVTFYFV56JZ7MXI6Z/bundle.json","state":"https://pith.science/pith/WICDBVJPGSVTFYFV56JZ7MXI6Z/state.json","well_known_bundle":"https://pith.science/.well-known/pith/WICDBVJPGSVTFYFV56JZ7MXI6Z/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:WICDBVJPGSVTFYFV56JZ7MXI6Z","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":"649e56ea8500ab3edfc8aa5ee3f2fe2edcc5365e4f0d8a8a68054d1aacb1611b","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-01-10T19:04:00Z","title_canon_sha256":"040523bacb11169d07d114f4b82f0261c04cf4a3a45d5b3e0d7f76797b0a9e32"},"schema_version":"1.0","source":{"id":"2401.05507","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2401.05507","created_at":"2026-07-05T07:54:35Z"},{"alias_kind":"arxiv_version","alias_value":"2401.05507v3","created_at":"2026-07-05T07:54:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.05507","created_at":"2026-07-05T07:54:35Z"},{"alias_kind":"pith_short_12","alias_value":"WICDBVJPGSVT","created_at":"2026-07-05T07:54:35Z"},{"alias_kind":"pith_short_16","alias_value":"WICDBVJPGSVTFYFV","created_at":"2026-07-05T07:54:35Z"},{"alias_kind":"pith_short_8","alias_value":"WICDBVJP","created_at":"2026-07-05T07:54:35Z"}],"graph_snapshots":[{"event_id":"sha256:863e8745a11881b96a2c511be3361dc73d484bb21c785f2e2172f28e99d07df1","target":"graph","created_at":"2026-07-05T07:54:35Z","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/2401.05507/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In this paper, we introduce InfiAgent-DABench, the first benchmark specifically designed to evaluate LLM-based agents on data analysis tasks. These tasks require agents to end-to-end solving complex tasks by interacting with an execution environment. This benchmark contains DAEval, a dataset consisting of 257 data analysis questions derived from 52 CSV files, and an agent framework which incorporates LLMs to serve as data analysis agents for both serving and evaluation. Since data analysis questions are often open-ended and hard to evaluate without human supervision, we adopt a format-promptin","authors_text":"Fei Wu, Guoyin Wang, Hongxia Yang, Jianbo Yuan, Jingjing Xu, Jing Su, Jiwei Li, Kun Kuang, Ming Zhu, Qianli Ma, Shuang Wei, Xueyu Hu, Xuwu Wang, Yang Yang, Yao Cheng, Ziwei Chai, Ziyu Zhao","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-01-10T19:04:00Z","title":"InfiAgent-DABench: Evaluating Agents on Data Analysis Tasks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.05507","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":"sha256:06435f973513f2e2c2d3af9493bd5fd1b0e5f99c20d88b7b5ead11e0211472b8","target":"record","created_at":"2026-07-05T07:54:35Z","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":"649e56ea8500ab3edfc8aa5ee3f2fe2edcc5365e4f0d8a8a68054d1aacb1611b","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-01-10T19:04:00Z","title_canon_sha256":"040523bacb11169d07d114f4b82f0261c04cf4a3a45d5b3e0d7f76797b0a9e32"},"schema_version":"1.0","source":{"id":"2401.05507","kind":"arxiv","version":3}},"canonical_sha256":"b20430d52f34ab32e0b5ef939fb2e8f6575307eeaf7cbce6b586ea71aa298a6f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b20430d52f34ab32e0b5ef939fb2e8f6575307eeaf7cbce6b586ea71aa298a6f","first_computed_at":"2026-07-05T07:54:35.317781Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:54:35.317781Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"8k0wj6sOoB02oTQmWFMBQc8bJe9KST6zG51SwqDMtFiVvYMc9vsB570wY4oYkrnkQNLEyF77kc1HklKCnloCCg==","signature_status":"signed_v1","signed_at":"2026-07-05T07:54:35.318276Z","signed_message":"canonical_sha256_bytes"},"source_id":"2401.05507","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:06435f973513f2e2c2d3af9493bd5fd1b0e5f99c20d88b7b5ead11e0211472b8","sha256:863e8745a11881b96a2c511be3361dc73d484bb21c785f2e2172f28e99d07df1"],"state_sha256":"566cd1317670019e6e32140b97494cb215e6aba6a41cee5c61a362a9d4b66deb"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tAGyNacIaWIdOtwPPHuouaQOaISYjovI634kshnrrbdxZmd1SbvufnJvMjdD1VONGqGKpYnldXk4ENCUxGGOCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T13:49:04.430232Z","bundle_sha256":"ad99fb4da1cb3939b200182a070f111e1414f89d48c20aa7522e51443ef08d49"}}