{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:ED6M5RPVEJ5QVVTAETO4QQETH2","short_pith_number":"pith:ED6M5RPV","schema_version":"1.0","canonical_sha256":"20fccec5f5227b0ad66024ddc840933ebb17926545f61b19278a4795f385e4bf","source":{"kind":"arxiv","id":"2401.06785","version":1},"attestation_state":"computed","paper":{"title":"Human-Instruction-Free LLM Self-Alignment with Limited Samples","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hongyi Guo, Jiaheng Wei, Wei Shen, Xiaoying Zhang, Yang Liu, Yuanshun Yao, Zhaoran Wang","submitted_at":"2024-01-06T14:00:12Z","abstract_excerpt":"Aligning large language models (LLMs) with human values is a vital task for LLM practitioners. Current alignment techniques have several limitations: (1) requiring a large amount of annotated data; (2) demanding heavy human involvement; (3) lacking a systematic mechanism to continuously improve. In this work, we study aligning LLMs to a new domain with limited samples (e.g. < 100). We propose an algorithm that can self-align LLMs iteratively without active human involvement. Unlike existing works, our algorithm relies on neither human-crafted instructions nor labeled rewards, significantly red"},"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":"2401.06785","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-01-06T14:00:12Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"05450a21b7db6f43197fdcb440c35d7bef2a2d36b0989b83a1eb29defb595123","abstract_canon_sha256":"a0a733c310aab809d90853194b5ec270aad3bbdca2325ac591b9e2814856f05c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:33:26.346338Z","signature_b64":"vYIqss6F+LKY+XZl6d/gS3dBbB8rHNeZ5Qx3ps0dOSIM5GQOorjRlAiVA0eGTNeuWdc9hqlICmDjfEuVqWp0CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"20fccec5f5227b0ad66024ddc840933ebb17926545f61b19278a4795f385e4bf","last_reissued_at":"2026-07-05T07:33:26.345835Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:33:26.345835Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Human-Instruction-Free LLM Self-Alignment with Limited Samples","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hongyi Guo, Jiaheng Wei, Wei Shen, Xiaoying Zhang, Yang Liu, Yuanshun Yao, Zhaoran Wang","submitted_at":"2024-01-06T14:00:12Z","abstract_excerpt":"Aligning large language models (LLMs) with human values is a vital task for LLM practitioners. Current alignment techniques have several limitations: (1) requiring a large amount of annotated data; (2) demanding heavy human involvement; (3) lacking a systematic mechanism to continuously improve. In this work, we study aligning LLMs to a new domain with limited samples (e.g. < 100). We propose an algorithm that can self-align LLMs iteratively without active human involvement. Unlike existing works, our algorithm relies on neither human-crafted instructions nor labeled rewards, significantly red"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.06785","kind":"arxiv","version":1},"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/2401.06785/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":"2401.06785","created_at":"2026-07-05T07:33:26.345896+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.06785v1","created_at":"2026-07-05T07:33:26.345896+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.06785","created_at":"2026-07-05T07:33:26.345896+00:00"},{"alias_kind":"pith_short_12","alias_value":"ED6M5RPVEJ5Q","created_at":"2026-07-05T07:33:26.345896+00:00"},{"alias_kind":"pith_short_16","alias_value":"ED6M5RPVEJ5QVVTA","created_at":"2026-07-05T07:33:26.345896+00:00"},{"alias_kind":"pith_short_8","alias_value":"ED6M5RPV","created_at":"2026-07-05T07:33:26.345896+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.07184","citing_title":"Structural Entropy Guided Agent for Detecting and Repairing Knowledge Deficiencies in LLMs","ref_index":9,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ED6M5RPVEJ5QVVTAETO4QQETH2","json":"https://pith.science/pith/ED6M5RPVEJ5QVVTAETO4QQETH2.json","graph_json":"https://pith.science/api/pith-number/ED6M5RPVEJ5QVVTAETO4QQETH2/graph.json","events_json":"https://pith.science/api/pith-number/ED6M5RPVEJ5QVVTAETO4QQETH2/events.json","paper":"https://pith.science/paper/ED6M5RPV"},"agent_actions":{"view_html":"https://pith.science/pith/ED6M5RPVEJ5QVVTAETO4QQETH2","download_json":"https://pith.science/pith/ED6M5RPVEJ5QVVTAETO4QQETH2.json","view_paper":"https://pith.science/paper/ED6M5RPV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.06785&json=true","fetch_graph":"https://pith.science/api/pith-number/ED6M5RPVEJ5QVVTAETO4QQETH2/graph.json","fetch_events":"https://pith.science/api/pith-number/ED6M5RPVEJ5QVVTAETO4QQETH2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ED6M5RPVEJ5QVVTAETO4QQETH2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ED6M5RPVEJ5QVVTAETO4QQETH2/action/storage_attestation","attest_author":"https://pith.science/pith/ED6M5RPVEJ5QVVTAETO4QQETH2/action/author_attestation","sign_citation":"https://pith.science/pith/ED6M5RPVEJ5QVVTAETO4QQETH2/action/citation_signature","submit_replication":"https://pith.science/pith/ED6M5RPVEJ5QVVTAETO4QQETH2/action/replication_record"}},"created_at":"2026-07-05T07:33:26.345896+00:00","updated_at":"2026-07-05T07:33:26.345896+00:00"}