{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:YUINIT4XEEV75LAI3ZFL67EMXG","short_pith_number":"pith:YUINIT4X","canonical_record":{"source":{"id":"2403.05307","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2024-03-08T13:34:20Z","cross_cats_sorted":[],"title_canon_sha256":"f39981cc845e08e2ceac5ebe954656df6e3501276a9188f000a466a8f3f65c66","abstract_canon_sha256":"c965dd5db88720a1e94534a9c5b9cb29cf11b02f78b45ed12c60f9a4271d4d2a"},"schema_version":"1.0"},"canonical_sha256":"c510d44f97212bfeac08de4abf7c8cb9a8e72ec5777e9bf7da4fb3412df4422a","source":{"kind":"arxiv","id":"2403.05307","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.05307","created_at":"2026-07-05T07:53:53Z"},{"alias_kind":"arxiv_version","alias_value":"2403.05307v1","created_at":"2026-07-05T07:53:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.05307","created_at":"2026-07-05T07:53:53Z"},{"alias_kind":"pith_short_12","alias_value":"YUINIT4XEEV7","created_at":"2026-07-05T07:53:53Z"},{"alias_kind":"pith_short_16","alias_value":"YUINIT4XEEV75LAI","created_at":"2026-07-05T07:53:53Z"},{"alias_kind":"pith_short_8","alias_value":"YUINIT4X","created_at":"2026-07-05T07:53:53Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:YUINIT4XEEV75LAI3ZFL67EMXG","target":"record","payload":{"canonical_record":{"source":{"id":"2403.05307","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2024-03-08T13:34:20Z","cross_cats_sorted":[],"title_canon_sha256":"f39981cc845e08e2ceac5ebe954656df6e3501276a9188f000a466a8f3f65c66","abstract_canon_sha256":"c965dd5db88720a1e94534a9c5b9cb29cf11b02f78b45ed12c60f9a4271d4d2a"},"schema_version":"1.0"},"canonical_sha256":"c510d44f97212bfeac08de4abf7c8cb9a8e72ec5777e9bf7da4fb3412df4422a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:53:53.495774Z","signature_b64":"b8s9mrwCVQmueKt4UpT9c4bqD3UohjAtuWOZfcLhz/v8OFo5JvZdxA4PB13RoV9snsKMTQlo9VGdGU0kWIYdBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c510d44f97212bfeac08de4abf7c8cb9a8e72ec5777e9bf7da4fb3412df4422a","last_reissued_at":"2026-07-05T07:53:53.495353Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:53:53.495353Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2403.05307","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":"2026-07-05T07:53:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Ua6WTK09OuB/3E5hsDtQpRnqYmTuDdFLo/EoX9iCWcZLI7QH1YwURCNysKQGH4GDmDnMiT/ioRi2bEJFN07zDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-11T07:19:31.054176Z"},"content_sha256":"08081b63505cb8a29d9b89d53ff3ed58fe45f8cb82ae835cd98fefe805228be5","schema_version":"1.0","event_id":"sha256:08081b63505cb8a29d9b89d53ff3ed58fe45f8cb82ae835cd98fefe805228be5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:YUINIT4XEEV75LAI3ZFL67EMXG","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Tapilot-Crossing: Benchmarking and Evolving LLMs Towards Interactive Data Analysis Agents","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chenhao Ma, Ge Qu, Jian-Guang Lou, Jiayi Shi, Jinyang Li, Nan Huo, Reynold Cheng, Yan Gao, Yingxiu Zhao, Yurong Wu","submitted_at":"2024-03-08T13:34:20Z","abstract_excerpt":"Interactive Data Analysis, the collaboration between humans and LLM agents, enables real-time data exploration for informed decision-making. The challenges and costs of collecting realistic interactive logs for data analysis hinder the quantitative evaluation of Large Language Model (LLM) agents in this task. To mitigate this issue, we introduce Tapilot-Crossing, a new benchmark to evaluate LLM agents on interactive data analysis. Tapilot-Crossing contains 1024 interactions, covering 4 practical scenarios: Normal, Action, Private, and Private Action. Notably, Tapilot-Crossing is constructed by"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.05307","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/2403.05307/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":"2026-07-05T07:53:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"R9p9pRzY5DJQ4bL4aBgE5YXz+LS22x5BcZ7oq5gLvYeXc35whLtAy9UkfpPdxaPFBwxSd/ib7GIrCM2RpEWjAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-11T07:19:31.055255Z"},"content_sha256":"8b0fea84627ca2f76fc2d4b7f39e21d0dd802cadcd45db71861ab96875d5f13d","schema_version":"1.0","event_id":"sha256:8b0fea84627ca2f76fc2d4b7f39e21d0dd802cadcd45db71861ab96875d5f13d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/YUINIT4XEEV75LAI3ZFL67EMXG/bundle.json","state_url":"https://pith.science/pith/YUINIT4XEEV75LAI3ZFL67EMXG/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/YUINIT4XEEV75LAI3ZFL67EMXG/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":"2026-08-11T07:19:31Z","links":{"resolver":"https://pith.science/pith/YUINIT4XEEV75LAI3ZFL67EMXG","bundle":"https://pith.science/pith/YUINIT4XEEV75LAI3ZFL67EMXG/bundle.json","state":"https://pith.science/pith/YUINIT4XEEV75LAI3ZFL67EMXG/state.json","well_known_bundle":"https://pith.science/.well-known/pith/YUINIT4XEEV75LAI3ZFL67EMXG/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:YUINIT4XEEV75LAI3ZFL67EMXG","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":"c965dd5db88720a1e94534a9c5b9cb29cf11b02f78b45ed12c60f9a4271d4d2a","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2024-03-08T13:34:20Z","title_canon_sha256":"f39981cc845e08e2ceac5ebe954656df6e3501276a9188f000a466a8f3f65c66"},"schema_version":"1.0","source":{"id":"2403.05307","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.05307","created_at":"2026-07-05T07:53:53Z"},{"alias_kind":"arxiv_version","alias_value":"2403.05307v1","created_at":"2026-07-05T07:53:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.05307","created_at":"2026-07-05T07:53:53Z"},{"alias_kind":"pith_short_12","alias_value":"YUINIT4XEEV7","created_at":"2026-07-05T07:53:53Z"},{"alias_kind":"pith_short_16","alias_value":"YUINIT4XEEV75LAI","created_at":"2026-07-05T07:53:53Z"},{"alias_kind":"pith_short_8","alias_value":"YUINIT4X","created_at":"2026-07-05T07:53:53Z"}],"graph_snapshots":[{"event_id":"sha256:8b0fea84627ca2f76fc2d4b7f39e21d0dd802cadcd45db71861ab96875d5f13d","target":"graph","created_at":"2026-07-05T07:53:53Z","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/2403.05307/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Interactive Data Analysis, the collaboration between humans and LLM agents, enables real-time data exploration for informed decision-making. The challenges and costs of collecting realistic interactive logs for data analysis hinder the quantitative evaluation of Large Language Model (LLM) agents in this task. To mitigate this issue, we introduce Tapilot-Crossing, a new benchmark to evaluate LLM agents on interactive data analysis. Tapilot-Crossing contains 1024 interactions, covering 4 practical scenarios: Normal, Action, Private, and Private Action. Notably, Tapilot-Crossing is constructed by","authors_text":"Chenhao Ma, Ge Qu, Jian-Guang Lou, Jiayi Shi, Jinyang Li, Nan Huo, Reynold Cheng, Yan Gao, Yingxiu Zhao, Yurong Wu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2024-03-08T13:34:20Z","title":"Tapilot-Crossing: Benchmarking and Evolving LLMs Towards Interactive Data Analysis Agents"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.05307","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":"sha256:08081b63505cb8a29d9b89d53ff3ed58fe45f8cb82ae835cd98fefe805228be5","target":"record","created_at":"2026-07-05T07:53:53Z","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":"c965dd5db88720a1e94534a9c5b9cb29cf11b02f78b45ed12c60f9a4271d4d2a","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2024-03-08T13:34:20Z","title_canon_sha256":"f39981cc845e08e2ceac5ebe954656df6e3501276a9188f000a466a8f3f65c66"},"schema_version":"1.0","source":{"id":"2403.05307","kind":"arxiv","version":1}},"canonical_sha256":"c510d44f97212bfeac08de4abf7c8cb9a8e72ec5777e9bf7da4fb3412df4422a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c510d44f97212bfeac08de4abf7c8cb9a8e72ec5777e9bf7da4fb3412df4422a","first_computed_at":"2026-07-05T07:53:53.495353Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:53:53.495353Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"b8s9mrwCVQmueKt4UpT9c4bqD3UohjAtuWOZfcLhz/v8OFo5JvZdxA4PB13RoV9snsKMTQlo9VGdGU0kWIYdBA==","signature_status":"signed_v1","signed_at":"2026-07-05T07:53:53.495774Z","signed_message":"canonical_sha256_bytes"},"source_id":"2403.05307","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:08081b63505cb8a29d9b89d53ff3ed58fe45f8cb82ae835cd98fefe805228be5","sha256:8b0fea84627ca2f76fc2d4b7f39e21d0dd802cadcd45db71861ab96875d5f13d"],"state_sha256":"b902deba488370079f66d28ff0e0011b5cbe6793dd417ae732a938994419bed8"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Bfm1xeSAyX/kz1kFVhfb5AHGjSaAX25VDrDBW6dzkRgaPokR6ZORWW27kmXsoIcf6gFS+YPND1MQKkPc+c5CCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-11T07:19:31.064025Z","bundle_sha256":"16f3f737b2f817558f18800d7e049042aba7f7d17e0810fdf5a422cd833470d6"}}