{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:RHK6GOL43XFUCXRRQPJDF2UQNK","short_pith_number":"pith:RHK6GOL4","schema_version":"1.0","canonical_sha256":"89d5e3397cddcb415e3183d232ea906a960c93fb48d0b4bb26c2cb4502613c2c","source":{"kind":"arxiv","id":"2405.03553","version":3},"attestation_state":"computed","paper":{"title":"AlphaMath Almost Zero: Process Supervision without Process","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chengxi Li, Guoxin Chen, Kai Fan, Minpeng Liao","submitted_at":"2024-05-06T15:20:30Z","abstract_excerpt":"Although recent advancements in large language models (LLMs) have significantly improved their performance on various tasks, they still face challenges with complex and symbolic multi-step reasoning, particularly in mathematical reasoning. To bolster the mathematical reasoning capabilities of LLMs, most existing efforts concentrate on seeking assistance from either domain experts or GPT-4 for high-quality process-supervised data, which is not only expensive but also labor-intensive. In our study, we propose an innovative framework, AlphaMath, that bypasses the need for process annotations (fro"},"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":"2405.03553","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-06T15:20:30Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f8f88ba37dd307a5e1ab58880e40b930bdf167c23c47b5d68ab99ead5b21ef47","abstract_canon_sha256":"0eaf5470755f714348a8baa286ec5a3f178bdfe5e5408fe201bbb855ade0ba7a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:12:28.230144Z","signature_b64":"l5P1b3G7a1nbhYsHHWa9rBK41lQ5xO+a6M3M+YK7jEQxzFv7ZzjrX+HW7HoXnuFQGFlYosHoocGHTwUWhKy/AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"89d5e3397cddcb415e3183d232ea906a960c93fb48d0b4bb26c2cb4502613c2c","last_reissued_at":"2026-07-05T09:12:28.229627Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:12:28.229627Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AlphaMath Almost Zero: Process Supervision without Process","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chengxi Li, Guoxin Chen, Kai Fan, Minpeng Liao","submitted_at":"2024-05-06T15:20:30Z","abstract_excerpt":"Although recent advancements in large language models (LLMs) have significantly improved their performance on various tasks, they still face challenges with complex and symbolic multi-step reasoning, particularly in mathematical reasoning. To bolster the mathematical reasoning capabilities of LLMs, most existing efforts concentrate on seeking assistance from either domain experts or GPT-4 for high-quality process-supervised data, which is not only expensive but also labor-intensive. In our study, we propose an innovative framework, AlphaMath, that bypasses the need for process annotations (fro"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.03553","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/2405.03553/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":"2405.03553","created_at":"2026-07-05T09:12:28.229690+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.03553v3","created_at":"2026-07-05T09:12:28.229690+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.03553","created_at":"2026-07-05T09:12:28.229690+00:00"},{"alias_kind":"pith_short_12","alias_value":"RHK6GOL43XFU","created_at":"2026-07-05T09:12:28.229690+00:00"},{"alias_kind":"pith_short_16","alias_value":"RHK6GOL43XFUCXRR","created_at":"2026-07-05T09:12:28.229690+00:00"},{"alias_kind":"pith_short_8","alias_value":"RHK6GOL4","created_at":"2026-07-05T09:12:28.229690+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":11,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21740","citing_title":"Training the Orchestrator: A Supervised Approach to End-to-End PDDL Planning with LLM Agents","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10568","citing_title":"VeriSpace: Spatially Grounded Action Verification for Vision-Language-Action Models","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05464","citing_title":"Step-by-Step Optimization-like Reasoning in LLMs over Expanding Search Spaces","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2406.18629","citing_title":"Step-DPO: Step-wise Preference Optimization for Long-chain Reasoning of LLMs","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2509.02547","citing_title":"The Landscape of Agentic Reinforcement Learning for LLMs: A Survey","ref_index":199,"is_internal_anchor":false},{"citing_arxiv_id":"2501.07301","citing_title":"The Lessons of Developing Process Reward Models in Mathematical Reasoning","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2501.09686","citing_title":"Towards Large Reasoning Models: A Survey of Reinforced Reasoning with Large Language Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2507.21046","citing_title":"A Survey of Self-Evolving Agents: What, When, How, and Where to Evolve on the Path to Artificial Super Intelligence","ref_index":233,"is_internal_anchor":false},{"citing_arxiv_id":"2502.17419","citing_title":"From System 1 to System 2: A Survey of Reasoning Large Language Models","ref_index":207,"is_internal_anchor":false},{"citing_arxiv_id":"2501.17161","citing_title":"SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10660","citing_title":"Efficient Process Reward Modeling via Contrastive Mutual Information","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RHK6GOL43XFUCXRRQPJDF2UQNK","json":"https://pith.science/pith/RHK6GOL43XFUCXRRQPJDF2UQNK.json","graph_json":"https://pith.science/api/pith-number/RHK6GOL43XFUCXRRQPJDF2UQNK/graph.json","events_json":"https://pith.science/api/pith-number/RHK6GOL43XFUCXRRQPJDF2UQNK/events.json","paper":"https://pith.science/paper/RHK6GOL4"},"agent_actions":{"view_html":"https://pith.science/pith/RHK6GOL43XFUCXRRQPJDF2UQNK","download_json":"https://pith.science/pith/RHK6GOL43XFUCXRRQPJDF2UQNK.json","view_paper":"https://pith.science/paper/RHK6GOL4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.03553&json=true","fetch_graph":"https://pith.science/api/pith-number/RHK6GOL43XFUCXRRQPJDF2UQNK/graph.json","fetch_events":"https://pith.science/api/pith-number/RHK6GOL43XFUCXRRQPJDF2UQNK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RHK6GOL43XFUCXRRQPJDF2UQNK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RHK6GOL43XFUCXRRQPJDF2UQNK/action/storage_attestation","attest_author":"https://pith.science/pith/RHK6GOL43XFUCXRRQPJDF2UQNK/action/author_attestation","sign_citation":"https://pith.science/pith/RHK6GOL43XFUCXRRQPJDF2UQNK/action/citation_signature","submit_replication":"https://pith.science/pith/RHK6GOL43XFUCXRRQPJDF2UQNK/action/replication_record"}},"created_at":"2026-07-05T09:12:28.229690+00:00","updated_at":"2026-07-05T09:12:28.229690+00:00"}