{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:HOFSSDHI4T6GDO5ZL3MVBIWUPB","short_pith_number":"pith:HOFSSDHI","schema_version":"1.0","canonical_sha256":"3b8b290ce8e4fc61bbb95ed950a2d4786ac0a6520c24042d79dd4faad22084fe","source":{"kind":"arxiv","id":"2402.01694","version":1},"attestation_state":"computed","paper":{"title":"ARGS: Alignment as Reward-Guided Search","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Jirayu Burapacheep, Maxim Khanov, Yixuan Li","submitted_at":"2024-01-23T23:42:41Z","abstract_excerpt":"Aligning large language models with human objectives is paramount, yet common approaches including RLHF suffer from unstable and resource-intensive training. In response to this challenge, we introduce ARGS, Alignment as Reward-Guided Search, a novel framework that integrates alignment into the decoding process, eliminating the need for expensive RL training. By adjusting the model's probabilistic predictions using a reward signal, ARGS generates texts with semantic diversity while being aligned with human preferences, offering a promising and flexible solution for aligning language models. No"},"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":"2402.01694","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-23T23:42:41Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"e01c13d50b98177b1e84e909c6eae764cc54f3d1ef63b6c1d580eda5f58f4207","abstract_canon_sha256":"22e2f2f28555172501498c0e00d3afaa624f4c949e78b48fd1e5239a018d439d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:41:05.197634Z","signature_b64":"6kA889uYrB1Vnboce6X7WYQ9GnymYwowXONLP9PacS+92rahbBa6joIbUOgXrp/TCM9TH7SITYuOdFCWaHHpDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3b8b290ce8e4fc61bbb95ed950a2d4786ac0a6520c24042d79dd4faad22084fe","last_reissued_at":"2026-07-05T07:41:05.197191Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:41:05.197191Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ARGS: Alignment as Reward-Guided Search","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Jirayu Burapacheep, Maxim Khanov, Yixuan Li","submitted_at":"2024-01-23T23:42:41Z","abstract_excerpt":"Aligning large language models with human objectives is paramount, yet common approaches including RLHF suffer from unstable and resource-intensive training. In response to this challenge, we introduce ARGS, Alignment as Reward-Guided Search, a novel framework that integrates alignment into the decoding process, eliminating the need for expensive RL training. By adjusting the model's probabilistic predictions using a reward signal, ARGS generates texts with semantic diversity while being aligned with human preferences, offering a promising and flexible solution for aligning language models. No"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.01694","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/2402.01694/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":"2402.01694","created_at":"2026-07-05T07:41:05.197261+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.01694v1","created_at":"2026-07-05T07:41:05.197261+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.01694","created_at":"2026-07-05T07:41:05.197261+00:00"},{"alias_kind":"pith_short_12","alias_value":"HOFSSDHI4T6G","created_at":"2026-07-05T07:41:05.197261+00:00"},{"alias_kind":"pith_short_16","alias_value":"HOFSSDHI4T6GDO5Z","created_at":"2026-07-05T07:41:05.197261+00:00"},{"alias_kind":"pith_short_8","alias_value":"HOFSSDHI","created_at":"2026-07-05T07:41:05.197261+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.04775","citing_title":"Activation Steering of Video Generation Models via Reduced-Order Linear Optimal Control","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2605.21851","citing_title":"OPPO: Bayesian Value Recursion for Token-Level Credit Assignment in LLM Reasoning","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.21851","citing_title":"OPPO: Bayesian Value Recursion for Token-Level Credit Assignment in LLM Reasoning","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20408","citing_title":"Spectral Souping: A Unified Framework for Online Preference Alignment","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10843","citing_title":"Training-Free Cultural Alignment of Large Language Models via Persona Disagreement","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2506.00166","citing_title":"Disentangled Safety Adapters Enable Efficient Guardrails and Flexible Inference-Time Alignment","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2507.06419","citing_title":"Teach a Reward Model to Correct Itself: Reward Guided Adversarial Failure Discovery for Robust Reward Modeling","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23543","citing_title":"Pref-CTRL: Preference Driven LLM Alignment using Representation Editing","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19018","citing_title":"Local Linearity of LLMs Enables Activation Steering via Model-Based Linear Optimal Control","ref_index":63,"is_internal_anchor":false},{"citing_arxiv_id":"2604.12479","citing_title":"Meet Dynamic Individual Preferences: Resolving Conflicting Human Value with Paired Fine-Tuning","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HOFSSDHI4T6GDO5ZL3MVBIWUPB","json":"https://pith.science/pith/HOFSSDHI4T6GDO5ZL3MVBIWUPB.json","graph_json":"https://pith.science/api/pith-number/HOFSSDHI4T6GDO5ZL3MVBIWUPB/graph.json","events_json":"https://pith.science/api/pith-number/HOFSSDHI4T6GDO5ZL3MVBIWUPB/events.json","paper":"https://pith.science/paper/HOFSSDHI"},"agent_actions":{"view_html":"https://pith.science/pith/HOFSSDHI4T6GDO5ZL3MVBIWUPB","download_json":"https://pith.science/pith/HOFSSDHI4T6GDO5ZL3MVBIWUPB.json","view_paper":"https://pith.science/paper/HOFSSDHI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.01694&json=true","fetch_graph":"https://pith.science/api/pith-number/HOFSSDHI4T6GDO5ZL3MVBIWUPB/graph.json","fetch_events":"https://pith.science/api/pith-number/HOFSSDHI4T6GDO5ZL3MVBIWUPB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HOFSSDHI4T6GDO5ZL3MVBIWUPB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HOFSSDHI4T6GDO5ZL3MVBIWUPB/action/storage_attestation","attest_author":"https://pith.science/pith/HOFSSDHI4T6GDO5ZL3MVBIWUPB/action/author_attestation","sign_citation":"https://pith.science/pith/HOFSSDHI4T6GDO5ZL3MVBIWUPB/action/citation_signature","submit_replication":"https://pith.science/pith/HOFSSDHI4T6GDO5ZL3MVBIWUPB/action/replication_record"}},"created_at":"2026-07-05T07:41:05.197261+00:00","updated_at":"2026-07-05T07:41:05.197261+00:00"}