{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:VDC7QLT4DFXPKKZLTW53MBLC4Z","short_pith_number":"pith:VDC7QLT4","schema_version":"1.0","canonical_sha256":"a8c5f82e7c196ef52b2b9dbbb60562e6513f196efacce699d5f229bd6a2bd065","source":{"kind":"arxiv","id":"2309.13078","version":2},"attestation_state":"computed","paper":{"title":"LPML: LLM-Prompting Markup Language for Mathematical Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.PL"],"primary_cat":"cs.AI","authors_text":"Akiyoshi Sannai, Ryutaro Yamauchi, Sho Sonoda, Wataru Kumagai","submitted_at":"2023-09-21T02:46:20Z","abstract_excerpt":"In utilizing large language models (LLMs) for mathematical reasoning, addressing the errors in the reasoning and calculation present in the generated text by LLMs is a crucial challenge. In this paper, we propose a novel framework that integrates the Chain-of-Thought (CoT) method with an external tool (Python REPL). We discovered that by prompting LLMs to generate structured text in XML-like markup language, we could seamlessly integrate CoT and the external tool and control the undesired behaviors of LLMs. With our approach, LLMs can utilize Python computation to rectify errors within CoT. We"},"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":"2309.13078","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2023-09-21T02:46:20Z","cross_cats_sorted":["cs.LG","cs.PL"],"title_canon_sha256":"43bdc2cc39a8ec79c4ded1c76167cb89fda88150c2306ed0a4eaeced3f36d58b","abstract_canon_sha256":"ab98c673c6f8c226fb51d03cacd6116c8e80e98ae7cc7c1dbf9c37a8cd65fb4d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:59:35.199811Z","signature_b64":"0HA/bb11pJrHm/CUhtAK+TFJQaRPksMwr+0uJF5pKB09tC6S56kc6TXN0LZ07Qi94I9egSTS+6TmbvJriO09CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a8c5f82e7c196ef52b2b9dbbb60562e6513f196efacce699d5f229bd6a2bd065","last_reissued_at":"2026-07-05T06:59:35.199331Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:59:35.199331Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LPML: LLM-Prompting Markup Language for Mathematical Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.PL"],"primary_cat":"cs.AI","authors_text":"Akiyoshi Sannai, Ryutaro Yamauchi, Sho Sonoda, Wataru Kumagai","submitted_at":"2023-09-21T02:46:20Z","abstract_excerpt":"In utilizing large language models (LLMs) for mathematical reasoning, addressing the errors in the reasoning and calculation present in the generated text by LLMs is a crucial challenge. In this paper, we propose a novel framework that integrates the Chain-of-Thought (CoT) method with an external tool (Python REPL). We discovered that by prompting LLMs to generate structured text in XML-like markup language, we could seamlessly integrate CoT and the external tool and control the undesired behaviors of LLMs. With our approach, LLMs can utilize Python computation to rectify errors within CoT. We"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.13078","kind":"arxiv","version":2},"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/2309.13078/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":"2309.13078","created_at":"2026-07-05T06:59:35.199393+00:00"},{"alias_kind":"arxiv_version","alias_value":"2309.13078v2","created_at":"2026-07-05T06:59:35.199393+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.13078","created_at":"2026-07-05T06:59:35.199393+00:00"},{"alias_kind":"pith_short_12","alias_value":"VDC7QLT4DFXP","created_at":"2026-07-05T06:59:35.199393+00:00"},{"alias_kind":"pith_short_16","alias_value":"VDC7QLT4DFXPKKZL","created_at":"2026-07-05T06:59:35.199393+00:00"},{"alias_kind":"pith_short_8","alias_value":"VDC7QLT4","created_at":"2026-07-05T06:59:35.199393+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.23533","citing_title":"POTracker: Optimizing Large Language Models for Standard-Compliant Power Outage Report Generation","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2606.23533","citing_title":"POTracker: Optimizing Large Language Models for Standard-Compliant Power Outage Report Generation","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2505.04588","citing_title":"ZeroSearch: Incentivize the Search Capability of LLMs without Searching","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2505.04588","citing_title":"ZeroSearch: Incentivize the Search Capability of LLMs without Searching","ref_index":40,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VDC7QLT4DFXPKKZLTW53MBLC4Z","json":"https://pith.science/pith/VDC7QLT4DFXPKKZLTW53MBLC4Z.json","graph_json":"https://pith.science/api/pith-number/VDC7QLT4DFXPKKZLTW53MBLC4Z/graph.json","events_json":"https://pith.science/api/pith-number/VDC7QLT4DFXPKKZLTW53MBLC4Z/events.json","paper":"https://pith.science/paper/VDC7QLT4"},"agent_actions":{"view_html":"https://pith.science/pith/VDC7QLT4DFXPKKZLTW53MBLC4Z","download_json":"https://pith.science/pith/VDC7QLT4DFXPKKZLTW53MBLC4Z.json","view_paper":"https://pith.science/paper/VDC7QLT4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2309.13078&json=true","fetch_graph":"https://pith.science/api/pith-number/VDC7QLT4DFXPKKZLTW53MBLC4Z/graph.json","fetch_events":"https://pith.science/api/pith-number/VDC7QLT4DFXPKKZLTW53MBLC4Z/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VDC7QLT4DFXPKKZLTW53MBLC4Z/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VDC7QLT4DFXPKKZLTW53MBLC4Z/action/storage_attestation","attest_author":"https://pith.science/pith/VDC7QLT4DFXPKKZLTW53MBLC4Z/action/author_attestation","sign_citation":"https://pith.science/pith/VDC7QLT4DFXPKKZLTW53MBLC4Z/action/citation_signature","submit_replication":"https://pith.science/pith/VDC7QLT4DFXPKKZLTW53MBLC4Z/action/replication_record"}},"created_at":"2026-07-05T06:59:35.199393+00:00","updated_at":"2026-07-05T06:59:35.199393+00:00"}