{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:7VD45JVVNEXHVYABZXJYTLGLY2","short_pith_number":"pith:7VD45JVV","schema_version":"1.0","canonical_sha256":"fd47cea6b5692e7ae001cdd389accbc6934d4fd9714f08b7aa337404f119d629","source":{"kind":"arxiv","id":"2402.09320","version":1},"attestation_state":"computed","paper":{"title":"ICDPO: Effectively Borrowing Alignment Capability of Others via In-context Direct Preference Optimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Feifan Song, Houfeng Wang, Peiyi Wang, Xin Zhang, Yuxuan Fan","submitted_at":"2024-02-14T17:14:34Z","abstract_excerpt":"Large Language Models (LLMs) rely on Human Preference Alignment (HPA) to ensure the generation of safe content. Due to the heavy cost associated with fine-tuning, fine-tuning-free methods have emerged, typically modifying LLM decoding with external auxiliary methods. However, these methods do not essentially enhance the LLM itself. In this paper, we rethink the derivation procedures of DPO, based on which we conversely build an instant scorer using the states of the LLM before and after In-context Learning (ICL). Accordingly, we propose a novel approach called In-Context Direct Preference Opti"},"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":"2402.09320","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-14T17:14:34Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a305f5ec1527e85d5ad27ebe6e568b7a7b68a183f8bb1f06a049b7b63f6c5974","abstract_canon_sha256":"28984cf850ca890993fe747d0cf1d75332547b29788a1aca90d03d5f13f8af93"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:45:12.155676Z","signature_b64":"SF1Tj0LD3pZQg9W2v0NhGyot2oIoHxIVutzH137KMDWrE7e3B/YM7XSXUkf9pfGoDmdXGBnLSV8CToJ3YV+yAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fd47cea6b5692e7ae001cdd389accbc6934d4fd9714f08b7aa337404f119d629","last_reissued_at":"2026-07-05T07:45:12.155236Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:45:12.155236Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ICDPO: Effectively Borrowing Alignment Capability of Others via In-context Direct Preference Optimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Feifan Song, Houfeng Wang, Peiyi Wang, Xin Zhang, Yuxuan Fan","submitted_at":"2024-02-14T17:14:34Z","abstract_excerpt":"Large Language Models (LLMs) rely on Human Preference Alignment (HPA) to ensure the generation of safe content. Due to the heavy cost associated with fine-tuning, fine-tuning-free methods have emerged, typically modifying LLM decoding with external auxiliary methods. However, these methods do not essentially enhance the LLM itself. In this paper, we rethink the derivation procedures of DPO, based on which we conversely build an instant scorer using the states of the LLM before and after In-context Learning (ICL). Accordingly, we propose a novel approach called In-Context Direct Preference Opti"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.09320","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/2402.09320/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":"2402.09320","created_at":"2026-07-05T07:45:12.155293+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.09320v1","created_at":"2026-07-05T07:45:12.155293+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.09320","created_at":"2026-07-05T07:45:12.155293+00:00"},{"alias_kind":"pith_short_12","alias_value":"7VD45JVVNEXH","created_at":"2026-07-05T07:45:12.155293+00:00"},{"alias_kind":"pith_short_16","alias_value":"7VD45JVVNEXHVYAB","created_at":"2026-07-05T07:45:12.155293+00:00"},{"alias_kind":"pith_short_8","alias_value":"7VD45JVV","created_at":"2026-07-05T07:45:12.155293+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.08720","citing_title":"Departures from Standard Disk Predictions in Intensive Ground-Based Monitoring of Three AGN","ref_index":26,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7VD45JVVNEXHVYABZXJYTLGLY2","json":"https://pith.science/pith/7VD45JVVNEXHVYABZXJYTLGLY2.json","graph_json":"https://pith.science/api/pith-number/7VD45JVVNEXHVYABZXJYTLGLY2/graph.json","events_json":"https://pith.science/api/pith-number/7VD45JVVNEXHVYABZXJYTLGLY2/events.json","paper":"https://pith.science/paper/7VD45JVV"},"agent_actions":{"view_html":"https://pith.science/pith/7VD45JVVNEXHVYABZXJYTLGLY2","download_json":"https://pith.science/pith/7VD45JVVNEXHVYABZXJYTLGLY2.json","view_paper":"https://pith.science/paper/7VD45JVV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.09320&json=true","fetch_graph":"https://pith.science/api/pith-number/7VD45JVVNEXHVYABZXJYTLGLY2/graph.json","fetch_events":"https://pith.science/api/pith-number/7VD45JVVNEXHVYABZXJYTLGLY2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7VD45JVVNEXHVYABZXJYTLGLY2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7VD45JVVNEXHVYABZXJYTLGLY2/action/storage_attestation","attest_author":"https://pith.science/pith/7VD45JVVNEXHVYABZXJYTLGLY2/action/author_attestation","sign_citation":"https://pith.science/pith/7VD45JVVNEXHVYABZXJYTLGLY2/action/citation_signature","submit_replication":"https://pith.science/pith/7VD45JVVNEXHVYABZXJYTLGLY2/action/replication_record"}},"created_at":"2026-07-05T07:45:12.155293+00:00","updated_at":"2026-07-05T07:45:12.155293+00:00"}