{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:7XHN7ARINL6YNGG6C7VTLF75ID","short_pith_number":"pith:7XHN7ARI","schema_version":"1.0","canonical_sha256":"fdcedf82286afd8698de17eb3597fd40e747105a630ed6e5462f6a07fbe648de","source":{"kind":"arxiv","id":"2505.02666","version":2},"attestation_state":"computed","paper":{"title":"A Survey on Progress in LLM Alignment from the Perspective of Reward Design","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jian Yang, Mark Dras, Miaomiao Ji, Shoujin Wang, Usman Naseem, Yanqiu Wu, Zhibin Wu","submitted_at":"2025-05-05T14:15:02Z","abstract_excerpt":"Reward design plays a pivotal role in aligning large language models (LLMs) with human values, serving as the bridge between feedback signals and model optimization. This survey provides a structured organization of reward modeling and addresses three key aspects: mathematical formulation, construction practices, and interaction with optimization paradigms. Building on this, it develops a macro-level taxonomy that characterizes reward mechanisms along complementary dimensions, thereby offering both conceptual clarity and practical guidance for alignment research. The progression of LLM alignme"},"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":"2505.02666","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-05T14:15:02Z","cross_cats_sorted":[],"title_canon_sha256":"2c4b9bf7f5cf63db83e174637c6887b4ecfd1489749a990e562f2a6b5b418740","abstract_canon_sha256":"e78d313e3d859960b696a86422e60bd9355e907eaf0ae5c71ea945c81f569294"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:03:26.474955Z","signature_b64":"+76nXRb72ZYf/VodxEFXZTThmWsBSA+KYQOegXutgZuRd03BwN95FHDXQD+ML6RWJI567ptNaDWNbuX3WDbFDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fdcedf82286afd8698de17eb3597fd40e747105a630ed6e5462f6a07fbe648de","last_reissued_at":"2026-07-05T12:03:26.474453Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:03:26.474453Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Survey on Progress in LLM Alignment from the Perspective of Reward Design","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jian Yang, Mark Dras, Miaomiao Ji, Shoujin Wang, Usman Naseem, Yanqiu Wu, Zhibin Wu","submitted_at":"2025-05-05T14:15:02Z","abstract_excerpt":"Reward design plays a pivotal role in aligning large language models (LLMs) with human values, serving as the bridge between feedback signals and model optimization. This survey provides a structured organization of reward modeling and addresses three key aspects: mathematical formulation, construction practices, and interaction with optimization paradigms. Building on this, it develops a macro-level taxonomy that characterizes reward mechanisms along complementary dimensions, thereby offering both conceptual clarity and practical guidance for alignment research. The progression of LLM alignme"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.02666","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/2505.02666/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":"2505.02666","created_at":"2026-07-05T12:03:26.474517+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.02666v2","created_at":"2026-07-05T12:03:26.474517+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.02666","created_at":"2026-07-05T12:03:26.474517+00:00"},{"alias_kind":"pith_short_12","alias_value":"7XHN7ARINL6Y","created_at":"2026-07-05T12:03:26.474517+00:00"},{"alias_kind":"pith_short_16","alias_value":"7XHN7ARINL6YNGG6","created_at":"2026-07-05T12:03:26.474517+00:00"},{"alias_kind":"pith_short_8","alias_value":"7XHN7ARI","created_at":"2026-07-05T12:03:26.474517+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.17458","citing_title":"ClaHF: A Human Feedback-inspired Reinforcement Learning Framework for Improving Classification Tasks","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2604.01346","citing_title":"Safety, Security, and Cognitive Risks in World Models","ref_index":68,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7XHN7ARINL6YNGG6C7VTLF75ID","json":"https://pith.science/pith/7XHN7ARINL6YNGG6C7VTLF75ID.json","graph_json":"https://pith.science/api/pith-number/7XHN7ARINL6YNGG6C7VTLF75ID/graph.json","events_json":"https://pith.science/api/pith-number/7XHN7ARINL6YNGG6C7VTLF75ID/events.json","paper":"https://pith.science/paper/7XHN7ARI"},"agent_actions":{"view_html":"https://pith.science/pith/7XHN7ARINL6YNGG6C7VTLF75ID","download_json":"https://pith.science/pith/7XHN7ARINL6YNGG6C7VTLF75ID.json","view_paper":"https://pith.science/paper/7XHN7ARI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.02666&json=true","fetch_graph":"https://pith.science/api/pith-number/7XHN7ARINL6YNGG6C7VTLF75ID/graph.json","fetch_events":"https://pith.science/api/pith-number/7XHN7ARINL6YNGG6C7VTLF75ID/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7XHN7ARINL6YNGG6C7VTLF75ID/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7XHN7ARINL6YNGG6C7VTLF75ID/action/storage_attestation","attest_author":"https://pith.science/pith/7XHN7ARINL6YNGG6C7VTLF75ID/action/author_attestation","sign_citation":"https://pith.science/pith/7XHN7ARINL6YNGG6C7VTLF75ID/action/citation_signature","submit_replication":"https://pith.science/pith/7XHN7ARINL6YNGG6C7VTLF75ID/action/replication_record"}},"created_at":"2026-07-05T12:03:26.474517+00:00","updated_at":"2026-07-05T12:03:26.474517+00:00"}