{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:5K7DRPJSUWHOT5FPUYXD33BYGH","short_pith_number":"pith:5K7DRPJS","canonical_record":{"source":{"id":"2407.01725","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-01T18:58:22Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"c85b9ff9b63a9edf2ebc11fad6f250e179f751451c722ad0bfd4af892f715329","abstract_canon_sha256":"4251f808dbd3f28c729e87a8e9df2d26db2a1160a13da5f3fea310235ee9f799"},"schema_version":"1.0"},"canonical_sha256":"eabe38bd32a58ee9f4afa62e3dec3831e051b4671ca91218744eb95190f7c08b","source":{"kind":"arxiv","id":"2407.01725","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.01725","created_at":"2026-07-05T08:39:04Z"},{"alias_kind":"arxiv_version","alias_value":"2407.01725v1","created_at":"2026-07-05T08:39:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.01725","created_at":"2026-07-05T08:39:04Z"},{"alias_kind":"pith_short_12","alias_value":"5K7DRPJSUWHO","created_at":"2026-07-05T08:39:04Z"},{"alias_kind":"pith_short_16","alias_value":"5K7DRPJSUWHOT5FP","created_at":"2026-07-05T08:39:04Z"},{"alias_kind":"pith_short_8","alias_value":"5K7DRPJS","created_at":"2026-07-05T08:39:04Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:5K7DRPJSUWHOT5FPUYXD33BYGH","target":"record","payload":{"canonical_record":{"source":{"id":"2407.01725","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-01T18:58:22Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"c85b9ff9b63a9edf2ebc11fad6f250e179f751451c722ad0bfd4af892f715329","abstract_canon_sha256":"4251f808dbd3f28c729e87a8e9df2d26db2a1160a13da5f3fea310235ee9f799"},"schema_version":"1.0"},"canonical_sha256":"eabe38bd32a58ee9f4afa62e3dec3831e051b4671ca91218744eb95190f7c08b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:39:04.226962Z","signature_b64":"mM6hB0lDkn/WFLMBcuW0wd6U6haW/vhrEMMlD2Tde+VoE2r0rnfCUYH0/n3uGBfFhUoXpxCxMC+fYHsFOloRCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"eabe38bd32a58ee9f4afa62e3dec3831e051b4671ca91218744eb95190f7c08b","last_reissued_at":"2026-07-05T08:39:04.226513Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:39:04.226513Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2407.01725","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":"2026-07-05T08:39:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"dkjl/zdtwXAz+kOfJZc05qwqzSoMQtlsIK1Q276UiQjHkf4F46RiXUXdB+M8wbD2P0Qu29Yt92T1I+xhgJozBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T07:34:11.257505Z"},"content_sha256":"1c22877aa6db9f6979206514ac6f5f986b9af1cc243ac2f954e0cb42ebf4d89e","schema_version":"1.0","event_id":"sha256:1c22877aa6db9f6979206514ac6f5f986b9af1cc243ac2f954e0cb42ebf4d89e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:5K7DRPJSUWHOT5FPUYXD33BYGH","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"DiscoveryBench: Towards Data-Driven Discovery with Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Abhijeetsingh Meena, Aryan Prakhar, Ashish Sabharwal, Bhavana Dalvi Mishra, Bodhisattwa Prasad Majumder, Dhruv Agarwal, Harshit Surana, Peter Clark, Tirth Vora, Tushar Khot","submitted_at":"2024-07-01T18:58:22Z","abstract_excerpt":"Can the rapid advances in code generation, function calling, and data analysis using large language models (LLMs) help automate the search and verification of hypotheses purely from a set of provided datasets? To evaluate this question, we present DiscoveryBench, the first comprehensive benchmark that formalizes the multi-step process of data-driven discovery. The benchmark is designed to systematically assess current model capabilities in discovery tasks and provide a useful resource for improving them. Our benchmark contains 264 tasks collected across 6 diverse domains, such as sociology and"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.01725","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/2407.01725/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":"2026-07-05T08:39:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"E8Ka2PaWbkerbdu0o02St87sdRQFnhzmmlkjnoa+nu6iaY3dVsa4tAyyO2ii2nb1+UH1NF6i9mbTeNRFg1qYDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T07:34:11.257886Z"},"content_sha256":"138ef28c76b18d60c37af740ec62a6d847389870cd38801f7d2f27e127b083e1","schema_version":"1.0","event_id":"sha256:138ef28c76b18d60c37af740ec62a6d847389870cd38801f7d2f27e127b083e1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/5K7DRPJSUWHOT5FPUYXD33BYGH/bundle.json","state_url":"https://pith.science/pith/5K7DRPJSUWHOT5FPUYXD33BYGH/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/5K7DRPJSUWHOT5FPUYXD33BYGH/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":"2026-07-06T07:34:11Z","links":{"resolver":"https://pith.science/pith/5K7DRPJSUWHOT5FPUYXD33BYGH","bundle":"https://pith.science/pith/5K7DRPJSUWHOT5FPUYXD33BYGH/bundle.json","state":"https://pith.science/pith/5K7DRPJSUWHOT5FPUYXD33BYGH/state.json","well_known_bundle":"https://pith.science/.well-known/pith/5K7DRPJSUWHOT5FPUYXD33BYGH/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:5K7DRPJSUWHOT5FPUYXD33BYGH","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":"4251f808dbd3f28c729e87a8e9df2d26db2a1160a13da5f3fea310235ee9f799","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-01T18:58:22Z","title_canon_sha256":"c85b9ff9b63a9edf2ebc11fad6f250e179f751451c722ad0bfd4af892f715329"},"schema_version":"1.0","source":{"id":"2407.01725","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.01725","created_at":"2026-07-05T08:39:04Z"},{"alias_kind":"arxiv_version","alias_value":"2407.01725v1","created_at":"2026-07-05T08:39:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.01725","created_at":"2026-07-05T08:39:04Z"},{"alias_kind":"pith_short_12","alias_value":"5K7DRPJSUWHO","created_at":"2026-07-05T08:39:04Z"},{"alias_kind":"pith_short_16","alias_value":"5K7DRPJSUWHOT5FP","created_at":"2026-07-05T08:39:04Z"},{"alias_kind":"pith_short_8","alias_value":"5K7DRPJS","created_at":"2026-07-05T08:39:04Z"}],"graph_snapshots":[{"event_id":"sha256:138ef28c76b18d60c37af740ec62a6d847389870cd38801f7d2f27e127b083e1","target":"graph","created_at":"2026-07-05T08:39:04Z","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/2407.01725/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Can the rapid advances in code generation, function calling, and data analysis using large language models (LLMs) help automate the search and verification of hypotheses purely from a set of provided datasets? To evaluate this question, we present DiscoveryBench, the first comprehensive benchmark that formalizes the multi-step process of data-driven discovery. The benchmark is designed to systematically assess current model capabilities in discovery tasks and provide a useful resource for improving them. Our benchmark contains 264 tasks collected across 6 diverse domains, such as sociology and","authors_text":"Abhijeetsingh Meena, Aryan Prakhar, Ashish Sabharwal, Bhavana Dalvi Mishra, Bodhisattwa Prasad Majumder, Dhruv Agarwal, Harshit Surana, Peter Clark, Tirth Vora, Tushar Khot","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-01T18:58:22Z","title":"DiscoveryBench: Towards Data-Driven Discovery with Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.01725","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":"sha256:1c22877aa6db9f6979206514ac6f5f986b9af1cc243ac2f954e0cb42ebf4d89e","target":"record","created_at":"2026-07-05T08:39:04Z","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":"4251f808dbd3f28c729e87a8e9df2d26db2a1160a13da5f3fea310235ee9f799","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-01T18:58:22Z","title_canon_sha256":"c85b9ff9b63a9edf2ebc11fad6f250e179f751451c722ad0bfd4af892f715329"},"schema_version":"1.0","source":{"id":"2407.01725","kind":"arxiv","version":1}},"canonical_sha256":"eabe38bd32a58ee9f4afa62e3dec3831e051b4671ca91218744eb95190f7c08b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"eabe38bd32a58ee9f4afa62e3dec3831e051b4671ca91218744eb95190f7c08b","first_computed_at":"2026-07-05T08:39:04.226513Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:39:04.226513Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"mM6hB0lDkn/WFLMBcuW0wd6U6haW/vhrEMMlD2Tde+VoE2r0rnfCUYH0/n3uGBfFhUoXpxCxMC+fYHsFOloRCA==","signature_status":"signed_v1","signed_at":"2026-07-05T08:39:04.226962Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.01725","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1c22877aa6db9f6979206514ac6f5f986b9af1cc243ac2f954e0cb42ebf4d89e","sha256:138ef28c76b18d60c37af740ec62a6d847389870cd38801f7d2f27e127b083e1"],"state_sha256":"19e4e2fe283a2ce9f641daa5348b78cd39978504479c93946d09de1cf163f807"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"S8EiuJxpuAxdEBSbrfN1jiuswve0InpJN62euxabbvHwpZ88wmc6nGY2tx+1HGJz0qknC/DVCw4xVphTM1pSBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T07:34:11.259891Z","bundle_sha256":"acf8d72bd6080044c2cfaff70d7fde9a7298c5b86d710e2b22a92ac7e8eaf59d"}}