{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:IRV77J2ISPCJOFNS76PNVLEPBK","short_pith_number":"pith:IRV77J2I","schema_version":"1.0","canonical_sha256":"446bffa74893c49715b2ff9edaac8f0a8616e424feaf78052a2795c4c121227c","source":{"kind":"arxiv","id":"2404.12253","version":2},"attestation_state":"computed","paper":{"title":"Toward Self-Improvement of LLMs via Imagination, Searching, and Criticizing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Baolin Peng, Dian Yu, Dong Yu, Haitao Mi, Lifeng Jin, Linfeng Song, Ye Tian","submitted_at":"2024-04-18T15:21:34Z","abstract_excerpt":"Despite the impressive capabilities of Large Language Models (LLMs) on various tasks, they still struggle with scenarios that involves complex reasoning and planning. Recent work proposed advanced prompting techniques and the necessity of fine-tuning with high-quality data to augment LLMs' reasoning abilities. However, these approaches are inherently constrained by data availability and quality. In light of this, self-correction and self-learning emerge as viable solutions, employing strategies that allow LLMs to refine their outputs and learn from self-assessed rewards. Yet, the efficacy of L"},"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":"2404.12253","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-18T15:21:34Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"6380e221687b39aa968976794c80d96736317ca7afe3c8ca506f9dd3564b1581","abstract_canon_sha256":"06d18e3e0dfcd076e8cd1cddfa2e3b37fcab24fddd9e5492df35562c32abeef7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:46:46.387270Z","signature_b64":"8cW3MCmYyqIf8CCYIj2pRIT0m1i7RjpnOP5dNs2XLUYqNY2fxTehX1Lfla4Z1QC8r//jgyjjbTpAhkdFTZHEDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"446bffa74893c49715b2ff9edaac8f0a8616e424feaf78052a2795c4c121227c","last_reissued_at":"2026-07-05T09:46:46.386686Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:46:46.386686Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Toward Self-Improvement of LLMs via Imagination, Searching, and Criticizing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Baolin Peng, Dian Yu, Dong Yu, Haitao Mi, Lifeng Jin, Linfeng Song, Ye Tian","submitted_at":"2024-04-18T15:21:34Z","abstract_excerpt":"Despite the impressive capabilities of Large Language Models (LLMs) on various tasks, they still struggle with scenarios that involves complex reasoning and planning. Recent work proposed advanced prompting techniques and the necessity of fine-tuning with high-quality data to augment LLMs' reasoning abilities. However, these approaches are inherently constrained by data availability and quality. In light of this, self-correction and self-learning emerge as viable solutions, employing strategies that allow LLMs to refine their outputs and learn from self-assessed rewards. Yet, the efficacy of L"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.12253","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/2404.12253/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":"2404.12253","created_at":"2026-07-05T09:46:46.386754+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.12253v2","created_at":"2026-07-05T09:46:46.386754+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.12253","created_at":"2026-07-05T09:46:46.386754+00:00"},{"alias_kind":"pith_short_12","alias_value":"IRV77J2ISPCJ","created_at":"2026-07-05T09:46:46.386754+00:00"},{"alias_kind":"pith_short_16","alias_value":"IRV77J2ISPCJOFNS","created_at":"2026-07-05T09:46:46.386754+00:00"},{"alias_kind":"pith_short_8","alias_value":"IRV77J2I","created_at":"2026-07-05T09:46:46.386754+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.07492","citing_title":"Search, Fail, Recover: A Training Framework for Correction-Aware Reasoning","ref_index":18,"is_internal_anchor":true},{"citing_arxiv_id":"2606.11470","citing_title":"The Periodic Table of LLM Reasoning: A Structured Survey of Reasoning Paradigms, Methods, and Failure Modes","ref_index":234,"is_internal_anchor":false},{"citing_arxiv_id":"2406.06592","citing_title":"Improve Mathematical Reasoning in Language Models by Automated Process Supervision","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2412.21187","citing_title":"Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs","ref_index":68,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09315","citing_title":"Do Self-Evolving Agents Forget? Capability Degradation and Preservation in Lifelong LLM Agent Adaptation","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2407.21787","citing_title":"Large Language Monkeys: Scaling Inference Compute with Repeated Sampling","ref_index":56,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IRV77J2ISPCJOFNS76PNVLEPBK","json":"https://pith.science/pith/IRV77J2ISPCJOFNS76PNVLEPBK.json","graph_json":"https://pith.science/api/pith-number/IRV77J2ISPCJOFNS76PNVLEPBK/graph.json","events_json":"https://pith.science/api/pith-number/IRV77J2ISPCJOFNS76PNVLEPBK/events.json","paper":"https://pith.science/paper/IRV77J2I"},"agent_actions":{"view_html":"https://pith.science/pith/IRV77J2ISPCJOFNS76PNVLEPBK","download_json":"https://pith.science/pith/IRV77J2ISPCJOFNS76PNVLEPBK.json","view_paper":"https://pith.science/paper/IRV77J2I","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.12253&json=true","fetch_graph":"https://pith.science/api/pith-number/IRV77J2ISPCJOFNS76PNVLEPBK/graph.json","fetch_events":"https://pith.science/api/pith-number/IRV77J2ISPCJOFNS76PNVLEPBK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IRV77J2ISPCJOFNS76PNVLEPBK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IRV77J2ISPCJOFNS76PNVLEPBK/action/storage_attestation","attest_author":"https://pith.science/pith/IRV77J2ISPCJOFNS76PNVLEPBK/action/author_attestation","sign_citation":"https://pith.science/pith/IRV77J2ISPCJOFNS76PNVLEPBK/action/citation_signature","submit_replication":"https://pith.science/pith/IRV77J2ISPCJOFNS76PNVLEPBK/action/replication_record"}},"created_at":"2026-07-05T09:46:46.386754+00:00","updated_at":"2026-07-05T09:46:46.386754+00:00"}