{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:CY4QRXCXBIFD44N73HW772CNEB","short_pith_number":"pith:CY4QRXCX","schema_version":"1.0","canonical_sha256":"163908dc570a0a3e71bfd9edffe84d2060072d06821847eaa4ea51d8e735bb73","source":{"kind":"arxiv","id":"2210.07128","version":3},"attestation_state":"computed","paper":{"title":"Language Models of Code are Few-Shot Commonsense Learners","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Aman Madaan, Graham Neubig, Shuyan Zhou, Uri Alon, Yiming Yang","submitted_at":"2022-10-13T16:09:36Z","abstract_excerpt":"We address the general task of structured commonsense reasoning: given a natural language input, the goal is to generate a graph such as an event -- or a reasoning-graph. To employ large language models (LMs) for this task, existing approaches ``serialize'' the output graph as a flat list of nodes and edges. Although feasible, these serialized graphs strongly deviate from the natural language corpora that LMs were pre-trained on, hindering LMs from generating them correctly. In this paper, we show that when we instead frame structured commonsense reasoning tasks as code generation tasks, pre-t"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2210.07128","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-10-13T16:09:36Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"17f9e4aab8314b4d59fc8ab503ec8815a879cd9cea0340fc5937ba0aea64805a","abstract_canon_sha256":"44837ffd873686f8e3f2ddaf2c520669f6053266ed311fd4a54d0bafcf5fc9b4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:22:57.488534Z","signature_b64":"TNJBfXqKLNw5XYQtOrsAuQngrjBqOcS8jVkKvNbVj96f1zuBGP8TuUniVunDk04N6G4RCO/r6IuLH9zyvNJrDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"163908dc570a0a3e71bfd9edffe84d2060072d06821847eaa4ea51d8e735bb73","last_reissued_at":"2026-07-05T05:22:57.488063Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:22:57.488063Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Language Models of Code are Few-Shot Commonsense Learners","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Aman Madaan, Graham Neubig, Shuyan Zhou, Uri Alon, Yiming Yang","submitted_at":"2022-10-13T16:09:36Z","abstract_excerpt":"We address the general task of structured commonsense reasoning: given a natural language input, the goal is to generate a graph such as an event -- or a reasoning-graph. To employ large language models (LMs) for this task, existing approaches ``serialize'' the output graph as a flat list of nodes and edges. Although feasible, these serialized graphs strongly deviate from the natural language corpora that LMs were pre-trained on, hindering LMs from generating them correctly. In this paper, we show that when we instead frame structured commonsense reasoning tasks as code generation tasks, pre-t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.07128","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/2210.07128/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2210.07128","created_at":"2026-07-05T05:22:57.488125+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.07128v3","created_at":"2026-07-05T05:22:57.488125+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.07128","created_at":"2026-07-05T05:22:57.488125+00:00"},{"alias_kind":"pith_short_12","alias_value":"CY4QRXCXBIFD","created_at":"2026-07-05T05:22:57.488125+00:00"},{"alias_kind":"pith_short_16","alias_value":"CY4QRXCXBIFD44N7","created_at":"2026-07-05T05:22:57.488125+00:00"},{"alias_kind":"pith_short_8","alias_value":"CY4QRXCX","created_at":"2026-07-05T05:22:57.488125+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.17929","citing_title":"PreAct: Computer-Using Agents that Get Faster on Repeated Tasks","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2303.08128","citing_title":"ViperGPT: Visual Inference via Python Execution for Reasoning","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2211.10435","citing_title":"PAL: Program-aided Language Models","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CY4QRXCXBIFD44N73HW772CNEB","json":"https://pith.science/pith/CY4QRXCXBIFD44N73HW772CNEB.json","graph_json":"https://pith.science/api/pith-number/CY4QRXCXBIFD44N73HW772CNEB/graph.json","events_json":"https://pith.science/api/pith-number/CY4QRXCXBIFD44N73HW772CNEB/events.json","paper":"https://pith.science/paper/CY4QRXCX"},"agent_actions":{"view_html":"https://pith.science/pith/CY4QRXCXBIFD44N73HW772CNEB","download_json":"https://pith.science/pith/CY4QRXCXBIFD44N73HW772CNEB.json","view_paper":"https://pith.science/paper/CY4QRXCX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.07128&json=true","fetch_graph":"https://pith.science/api/pith-number/CY4QRXCXBIFD44N73HW772CNEB/graph.json","fetch_events":"https://pith.science/api/pith-number/CY4QRXCXBIFD44N73HW772CNEB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CY4QRXCXBIFD44N73HW772CNEB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CY4QRXCXBIFD44N73HW772CNEB/action/storage_attestation","attest_author":"https://pith.science/pith/CY4QRXCXBIFD44N73HW772CNEB/action/author_attestation","sign_citation":"https://pith.science/pith/CY4QRXCXBIFD44N73HW772CNEB/action/citation_signature","submit_replication":"https://pith.science/pith/CY4QRXCXBIFD44N73HW772CNEB/action/replication_record"}},"created_at":"2026-07-05T05:22:57.488125+00:00","updated_at":"2026-07-05T05:22:57.488125+00:00"}