{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:NA5YCIP46X6WRX75FU44EZSWTX","short_pith_number":"pith:NA5YCIP4","schema_version":"1.0","canonical_sha256":"683b8121fcf5fd68dffd2d39c266569dc16600a97ba30db3480add81ead715f0","source":{"kind":"arxiv","id":"2501.03226","version":3},"attestation_state":"computed","paper":{"title":"BoostStep: Boosting mathematical capability of Large Language Models via improved single-step reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Beichen Zhang, Dahua Lin, Haodong Duan, Jiaqi Wang, Pan Zhang, Xiaoyi Dong, Yuhang Cao, Yuhang Zang, Yuhong Liu","submitted_at":"2025-01-06T18:59:13Z","abstract_excerpt":"Large language models (LLMs) have demonstrated impressive ability in solving complex mathematical problems with multi-step reasoning and can be further enhanced with well-designed in-context learning (ICL) examples. However, this potential is often constrained by two major challenges in ICL: granularity mismatch and irrelevant information. We observe that while LLMs excel at decomposing mathematical problems, they often struggle with reasoning errors in fine-grained steps. Moreover, ICL examples retrieved at the question level may omit critical steps or even mislead the model with irrelevant d"},"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":"2501.03226","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-06T18:59:13Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"f90b7054c562c1bb87247ef213e141922d25caf6ffc9653537b8b1f9f019c0fe","abstract_canon_sha256":"401b43954ffd3bf2ecac891471a1903f0b083605fd7c3503341ac93a91fd011f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:15:24.398955Z","signature_b64":"XobGD6jRen6DS1//T3UiHoMcNGphIIwU/wxmQzweVlTOkPIz6cdV3AkTXsOo81agKKQ0h2lRKjWg4iuq/oiwBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"683b8121fcf5fd68dffd2d39c266569dc16600a97ba30db3480add81ead715f0","last_reissued_at":"2026-07-05T10:15:24.398458Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:15:24.398458Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"BoostStep: Boosting mathematical capability of Large Language Models via improved single-step reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Beichen Zhang, Dahua Lin, Haodong Duan, Jiaqi Wang, Pan Zhang, Xiaoyi Dong, Yuhang Cao, Yuhang Zang, Yuhong Liu","submitted_at":"2025-01-06T18:59:13Z","abstract_excerpt":"Large language models (LLMs) have demonstrated impressive ability in solving complex mathematical problems with multi-step reasoning and can be further enhanced with well-designed in-context learning (ICL) examples. However, this potential is often constrained by two major challenges in ICL: granularity mismatch and irrelevant information. We observe that while LLMs excel at decomposing mathematical problems, they often struggle with reasoning errors in fine-grained steps. Moreover, ICL examples retrieved at the question level may omit critical steps or even mislead the model with irrelevant d"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.03226","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/2501.03226/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":"2501.03226","created_at":"2026-07-05T10:15:24.398521+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.03226v3","created_at":"2026-07-05T10:15:24.398521+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.03226","created_at":"2026-07-05T10:15:24.398521+00:00"},{"alias_kind":"pith_short_12","alias_value":"NA5YCIP46X6W","created_at":"2026-07-05T10:15:24.398521+00:00"},{"alias_kind":"pith_short_16","alias_value":"NA5YCIP46X6WRX75","created_at":"2026-07-05T10:15:24.398521+00:00"},{"alias_kind":"pith_short_8","alias_value":"NA5YCIP4","created_at":"2026-07-05T10:15:24.398521+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2509.06284","citing_title":"From Implicit Exploration to Structured Reasoning: Leveraging Guideline and Refinement for LLMs","ref_index":14,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NA5YCIP46X6WRX75FU44EZSWTX","json":"https://pith.science/pith/NA5YCIP46X6WRX75FU44EZSWTX.json","graph_json":"https://pith.science/api/pith-number/NA5YCIP46X6WRX75FU44EZSWTX/graph.json","events_json":"https://pith.science/api/pith-number/NA5YCIP46X6WRX75FU44EZSWTX/events.json","paper":"https://pith.science/paper/NA5YCIP4"},"agent_actions":{"view_html":"https://pith.science/pith/NA5YCIP46X6WRX75FU44EZSWTX","download_json":"https://pith.science/pith/NA5YCIP46X6WRX75FU44EZSWTX.json","view_paper":"https://pith.science/paper/NA5YCIP4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.03226&json=true","fetch_graph":"https://pith.science/api/pith-number/NA5YCIP46X6WRX75FU44EZSWTX/graph.json","fetch_events":"https://pith.science/api/pith-number/NA5YCIP46X6WRX75FU44EZSWTX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NA5YCIP46X6WRX75FU44EZSWTX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NA5YCIP46X6WRX75FU44EZSWTX/action/storage_attestation","attest_author":"https://pith.science/pith/NA5YCIP46X6WRX75FU44EZSWTX/action/author_attestation","sign_citation":"https://pith.science/pith/NA5YCIP46X6WRX75FU44EZSWTX/action/citation_signature","submit_replication":"https://pith.science/pith/NA5YCIP46X6WRX75FU44EZSWTX/action/replication_record"}},"created_at":"2026-07-05T10:15:24.398521+00:00","updated_at":"2026-07-05T10:15:24.398521+00:00"}