{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:YYZDSVTLHRFFDPBDLOWEVOOIGQ","short_pith_number":"pith:YYZDSVTL","schema_version":"1.0","canonical_sha256":"c63239566b3c4a51bc235bac4ab9c834293c1b0f6ae3b964a8ccaf66cc688c12","source":{"kind":"arxiv","id":"2207.14502","version":4},"attestation_state":"computed","paper":{"title":"Language Models Can Teach Themselves to Program Better","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Adam Tauman Kalai, Matthew Bowers, Patrick Haluptzok","submitted_at":"2022-07-29T06:43:28Z","abstract_excerpt":"Recent Language Models (LMs) achieve breakthrough performance in code generation when trained on human-authored problems, even solving some competitive-programming problems. Self-play has proven useful in games such as Go, and thus it is natural to ask whether LMs can generate their own instructive programming problems to improve their performance. We show that it is possible for an LM to synthesize programming problems and solutions, which are filtered for correctness by a Python interpreter. The LM's performance is then seen to improve when it is fine-tuned on its own synthetic problems and "},"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":"2207.14502","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-07-29T06:43:28Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"512204258a8855d0a36d8136b6e40316e26dff54acc00474f95f62c52b5f2b12","abstract_canon_sha256":"7c1ac5ba45d82f775d44e010eebc0edb77d929695ae6c8fd9ab94e098ec85974"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:00:12.579776Z","signature_b64":"xcVo0/RRVc/Sf2Q4dO7KtWEr0DEZz68EI0f9n+WVAczT8LsexsZSoBlxR4XExYAQh86SHP63qvP8VF66ZQPqCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c63239566b3c4a51bc235bac4ab9c834293c1b0f6ae3b964a8ccaf66cc688c12","last_reissued_at":"2026-07-05T06:00:12.579399Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:00:12.579399Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Language Models Can Teach Themselves to Program Better","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Adam Tauman Kalai, Matthew Bowers, Patrick Haluptzok","submitted_at":"2022-07-29T06:43:28Z","abstract_excerpt":"Recent Language Models (LMs) achieve breakthrough performance in code generation when trained on human-authored problems, even solving some competitive-programming problems. Self-play has proven useful in games such as Go, and thus it is natural to ask whether LMs can generate their own instructive programming problems to improve their performance. We show that it is possible for an LM to synthesize programming problems and solutions, which are filtered for correctness by a Python interpreter. The LM's performance is then seen to improve when it is fine-tuned on its own synthetic problems and "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2207.14502","kind":"arxiv","version":4},"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/2207.14502/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":"2207.14502","created_at":"2026-07-05T06:00:12.579457+00:00"},{"alias_kind":"arxiv_version","alias_value":"2207.14502v4","created_at":"2026-07-05T06:00:12.579457+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2207.14502","created_at":"2026-07-05T06:00:12.579457+00:00"},{"alias_kind":"pith_short_12","alias_value":"YYZDSVTLHRFF","created_at":"2026-07-05T06:00:12.579457+00:00"},{"alias_kind":"pith_short_16","alias_value":"YYZDSVTLHRFFDPBD","created_at":"2026-07-05T06:00:12.579457+00:00"},{"alias_kind":"pith_short_8","alias_value":"YYZDSVTL","created_at":"2026-07-05T06:00:12.579457+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.02390","citing_title":"DecompRL: Solving Harder Problems by Learning Modular Code Generation","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2503.14434","citing_title":"LLM-FE: Automated Feature Engineering for Tabular Data with LLMs as Evolutionary Optimizers","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2509.20881","citing_title":"PseudoBridge: Pseudo Code as the Bridge for Better Semantic and Logic Alignment in Code Retrieval","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2512.18552","citing_title":"Toward Training Superintelligent Software Agents through Self-Play SWE-RL","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2510.25223","citing_title":"FELA: A Multi-Agent Evolutionary System for Feature Engineering of Industrial Event Log Data","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2406.10162","citing_title":"Sycophancy to Subterfuge: Investigating Reward-Tampering in Large Language Models","ref_index":167,"is_internal_anchor":false},{"citing_arxiv_id":"2403.07974","citing_title":"LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code","ref_index":40,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YYZDSVTLHRFFDPBDLOWEVOOIGQ","json":"https://pith.science/pith/YYZDSVTLHRFFDPBDLOWEVOOIGQ.json","graph_json":"https://pith.science/api/pith-number/YYZDSVTLHRFFDPBDLOWEVOOIGQ/graph.json","events_json":"https://pith.science/api/pith-number/YYZDSVTLHRFFDPBDLOWEVOOIGQ/events.json","paper":"https://pith.science/paper/YYZDSVTL"},"agent_actions":{"view_html":"https://pith.science/pith/YYZDSVTLHRFFDPBDLOWEVOOIGQ","download_json":"https://pith.science/pith/YYZDSVTLHRFFDPBDLOWEVOOIGQ.json","view_paper":"https://pith.science/paper/YYZDSVTL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2207.14502&json=true","fetch_graph":"https://pith.science/api/pith-number/YYZDSVTLHRFFDPBDLOWEVOOIGQ/graph.json","fetch_events":"https://pith.science/api/pith-number/YYZDSVTLHRFFDPBDLOWEVOOIGQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YYZDSVTLHRFFDPBDLOWEVOOIGQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YYZDSVTLHRFFDPBDLOWEVOOIGQ/action/storage_attestation","attest_author":"https://pith.science/pith/YYZDSVTLHRFFDPBDLOWEVOOIGQ/action/author_attestation","sign_citation":"https://pith.science/pith/YYZDSVTLHRFFDPBDLOWEVOOIGQ/action/citation_signature","submit_replication":"https://pith.science/pith/YYZDSVTLHRFFDPBDLOWEVOOIGQ/action/replication_record"}},"created_at":"2026-07-05T06:00:12.579457+00:00","updated_at":"2026-07-05T06:00:12.579457+00:00"}