{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:HJUHWMM55R5TTKQFS4RAO6SSXZ","short_pith_number":"pith:HJUHWMM5","schema_version":"1.0","canonical_sha256":"3a687b319dec7b39aa059722077a52be6a3fc8b92c3ceb98ab048bfab4d4e9f1","source":{"kind":"arxiv","id":"2509.05605","version":1},"attestation_state":"computed","paper":{"title":"Icon$^{2}$: Aligning Large Language Models Using Self-Synthetic Preference Data via Inherent Regulation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hongsen Huang, Hongxia Xu, Jiahe Chen, Jian Wu, Jintai Chen, Qian Shao, Qiyuan Chen, Ren Chuan, Renjie Hua","submitted_at":"2025-09-06T05:38:47Z","abstract_excerpt":"Large Language Models (LLMs) require high quality preference datasets to align with human preferences. However, conventional methods for constructing such datasets face significant challenges: reliance on pre-collected instructions often leads to distribution mismatches with target models, while the need for sampling multiple stochastic responses introduces substantial computational overhead. In this work, we explore a paradigm shift by leveraging inherent regulation of LLMs' representation space for efficient and tailored preference dataset construction, named Icon$^{2}$. Specifically, it fir"},"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":"2509.05605","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-09-06T05:38:47Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"b1e34b8d7f97707acc520355abc71ead3cfd2bde12505cdf61ccdbd75455eae1","abstract_canon_sha256":"9f45f98c0a41dc4f6a7c97b12b962142948068ba323f73edaf6da1f27d22a7f2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:06:20.926174Z","signature_b64":"pj7MfnCf1yEURwD15wya0KvDxegGu+wlkhFfPD6ba4tsXxvOY0uF8KeqF9YkbEAlkbdQOi+gDHPfLpxz3+cEDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3a687b319dec7b39aa059722077a52be6a3fc8b92c3ceb98ab048bfab4d4e9f1","last_reissued_at":"2026-07-05T12:06:20.925615Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:06:20.925615Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Icon$^{2}$: Aligning Large Language Models Using Self-Synthetic Preference Data via Inherent Regulation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hongsen Huang, Hongxia Xu, Jiahe Chen, Jian Wu, Jintai Chen, Qian Shao, Qiyuan Chen, Ren Chuan, Renjie Hua","submitted_at":"2025-09-06T05:38:47Z","abstract_excerpt":"Large Language Models (LLMs) require high quality preference datasets to align with human preferences. However, conventional methods for constructing such datasets face significant challenges: reliance on pre-collected instructions often leads to distribution mismatches with target models, while the need for sampling multiple stochastic responses introduces substantial computational overhead. In this work, we explore a paradigm shift by leveraging inherent regulation of LLMs' representation space for efficient and tailored preference dataset construction, named Icon$^{2}$. Specifically, it fir"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.05605","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/2509.05605/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":"2509.05605","created_at":"2026-07-05T12:06:20.925691+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.05605v1","created_at":"2026-07-05T12:06:20.925691+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.05605","created_at":"2026-07-05T12:06:20.925691+00:00"},{"alias_kind":"pith_short_12","alias_value":"HJUHWMM55R5T","created_at":"2026-07-05T12:06:20.925691+00:00"},{"alias_kind":"pith_short_16","alias_value":"HJUHWMM55R5TTKQF","created_at":"2026-07-05T12:06:20.925691+00:00"},{"alias_kind":"pith_short_8","alias_value":"HJUHWMM5","created_at":"2026-07-05T12:06:20.925691+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HJUHWMM55R5TTKQFS4RAO6SSXZ","json":"https://pith.science/pith/HJUHWMM55R5TTKQFS4RAO6SSXZ.json","graph_json":"https://pith.science/api/pith-number/HJUHWMM55R5TTKQFS4RAO6SSXZ/graph.json","events_json":"https://pith.science/api/pith-number/HJUHWMM55R5TTKQFS4RAO6SSXZ/events.json","paper":"https://pith.science/paper/HJUHWMM5"},"agent_actions":{"view_html":"https://pith.science/pith/HJUHWMM55R5TTKQFS4RAO6SSXZ","download_json":"https://pith.science/pith/HJUHWMM55R5TTKQFS4RAO6SSXZ.json","view_paper":"https://pith.science/paper/HJUHWMM5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.05605&json=true","fetch_graph":"https://pith.science/api/pith-number/HJUHWMM55R5TTKQFS4RAO6SSXZ/graph.json","fetch_events":"https://pith.science/api/pith-number/HJUHWMM55R5TTKQFS4RAO6SSXZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HJUHWMM55R5TTKQFS4RAO6SSXZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HJUHWMM55R5TTKQFS4RAO6SSXZ/action/storage_attestation","attest_author":"https://pith.science/pith/HJUHWMM55R5TTKQFS4RAO6SSXZ/action/author_attestation","sign_citation":"https://pith.science/pith/HJUHWMM55R5TTKQFS4RAO6SSXZ/action/citation_signature","submit_replication":"https://pith.science/pith/HJUHWMM55R5TTKQFS4RAO6SSXZ/action/replication_record"}},"created_at":"2026-07-05T12:06:20.925691+00:00","updated_at":"2026-07-05T12:06:20.925691+00:00"}