{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:63NRELHFFN5IYRD4CHKJSYLPIL","short_pith_number":"pith:63NRELHF","schema_version":"1.0","canonical_sha256":"f6db122ce52b7a8c447c11d499616f42fbb5cdaaa2db7b6b45e455121da19756","source":{"kind":"arxiv","id":"2605.13875","version":1},"attestation_state":"computed","paper":{"title":"Common-agency Games for Multi-Objective Test-Time Alignment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"CAGE treats multiple conflicting alignment goals as strategic principals bidding token incentives to produce an equilibrium LLM policy at inference time.","cross_cats":[],"primary_cat":"cs.GT","authors_text":"Baiting Chen, Rui Yu, Tong Zhu, Xiaowu Dai","submitted_at":"2026-05-08T06:56:35Z","abstract_excerpt":"Aligning large language models (LLMs) with human preferences is inherently multi-objective: different users and evaluation criteria impose heterogeneous and often conflicting requirements on model outputs. We propose CAGE (Common-Agency Games for Alignment), a training-free, game-theoretic framework for multi-objective test-time alignment. CAGE models alignment objectives as strategic principals that allocate token-level incentives to a shared LLM, inducing an equilibrium policy that captures the joint effect of competing objectives. We develop an efficient algorithm based on equilibrium probl"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2605.13875","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.GT","submitted_at":"2026-05-08T06:56:35Z","cross_cats_sorted":[],"title_canon_sha256":"ec01d02e8ddaea2090a0c9969382906a7bd8f682e5499ce7de30b309fee65ba3","abstract_canon_sha256":"0d7b2fceed870fc9915bb543fe1bc4aad71a2254498f74720408394b6965a165"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:19.263731Z","signature_b64":"WzcrMuPSQSdee8hb6nNyW8rcTfw7z55U7gkmVVTJKWiizLKZ2B75LUGODvyhrGPG5dlzFtE7vWJh3HPZoiQvBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f6db122ce52b7a8c447c11d499616f42fbb5cdaaa2db7b6b45e455121da19756","last_reissued_at":"2026-05-17T23:39:19.262842Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:19.262842Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Common-agency Games for Multi-Objective Test-Time Alignment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"CAGE treats multiple conflicting alignment goals as strategic principals bidding token incentives to produce an equilibrium LLM policy at inference time.","cross_cats":[],"primary_cat":"cs.GT","authors_text":"Baiting Chen, Rui Yu, Tong Zhu, Xiaowu Dai","submitted_at":"2026-05-08T06:56:35Z","abstract_excerpt":"Aligning large language models (LLMs) with human preferences is inherently multi-objective: different users and evaluation criteria impose heterogeneous and often conflicting requirements on model outputs. We propose CAGE (Common-Agency Games for Alignment), a training-free, game-theoretic framework for multi-objective test-time alignment. CAGE models alignment objectives as strategic principals that allocate token-level incentives to a shared LLM, inducing an equilibrium policy that captures the joint effect of competing objectives. We develop an efficient algorithm based on equilibrium probl"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"CAGE enables flexible and fine-grained trade-offs across objectives at inference time, consistently outperforming existing test-time alignment methods while requiring no retraining. It further supports weak-to-strong generalization, making multi-objective alignment practical in resource-constrained settings.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That modeling heterogeneous objectives as strategic principals allocating token-level incentives produces an equilibrium policy whose joint effect meaningfully captures real user preferences, and that the EPEC-based algorithm reliably computes this equilibrium with the claimed existence, uniqueness, convergence, and stability guarantees.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"CAGE uses common-agency games and an EPEC algorithm to compute equilibrium policies that balance multiple conflicting objectives for test-time LLM alignment.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"CAGE treats multiple conflicting alignment goals as strategic principals bidding token incentives to produce an equilibrium LLM policy at inference time.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"ba3aeb72703cb6e24891c5dae8672767889a68094fa699046c184755d7700103"},"source":{"id":"2605.13875","kind":"arxiv","version":1},"verdict":{"id":"1c5a5ac2-e479-4d34-af9e-4253368271a2","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T06:11:14.544441Z","strongest_claim":"CAGE enables flexible and fine-grained trade-offs across objectives at inference time, consistently outperforming existing test-time alignment methods while requiring no retraining. It further supports weak-to-strong generalization, making multi-objective alignment practical in resource-constrained settings.","one_line_summary":"CAGE uses common-agency games and an EPEC algorithm to compute equilibrium policies that balance multiple conflicting objectives for test-time LLM alignment.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That modeling heterogeneous objectives as strategic principals allocating token-level incentives produces an equilibrium policy whose joint effect meaningfully captures real user preferences, and that the EPEC-based algorithm reliably computes this equilibrium with the claimed existence, uniqueness, convergence, and stability guarantees.","pith_extraction_headline":"CAGE treats multiple conflicting alignment goals as strategic principals bidding token incentives to produce an equilibrium LLM policy at inference time."},"references":{"count":230,"sample":[{"doi":"","year":2024,"title":"TinyLlama: An Open-Source Small Language Model , author=. 2024 , eprint=","work_id":"4b2a85e5-7934-4e57-ba32-e7ac0a26d388","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"International Conference on Learning Representations , year=","work_id":"7d66a0d4-a1e1-4771-b0cd-ea8c424cb4bd","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"The Twelfth International Conference on Learning Representations , year=","work_id":"44d6bf99-5902-43da-b8fd-83297976c6e0","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"C hat D ev: Communicative agents for software development","work_id":"166b34d2-9477-4b53-b493-bcce91575006","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"Encouraging divergent thinking in large language models through multi-agent debate","work_id":"5e657239-8ab5-4c3c-ac58-63b4c9d2e420","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":230,"snapshot_sha256":"0d6a0ec0b4c4362b2aaaab7e88e2dcbc27332505742a1851b0d20041ea2a4151","internal_anchors":19},"formal_canon":{"evidence_count":2,"snapshot_sha256":"aebd8280c17417da790c7bdeedd992a81e4524c1d32020d083c0fcbfb65b478e"},"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":"2605.13875","created_at":"2026-05-17T23:39:19.263000+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.13875v1","created_at":"2026-05-17T23:39:19.263000+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13875","created_at":"2026-05-17T23:39:19.263000+00:00"},{"alias_kind":"pith_short_12","alias_value":"63NRELHFFN5I","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"63NRELHFFN5IYRD4","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"63NRELHF","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/63NRELHFFN5IYRD4CHKJSYLPIL","json":"https://pith.science/pith/63NRELHFFN5IYRD4CHKJSYLPIL.json","graph_json":"https://pith.science/api/pith-number/63NRELHFFN5IYRD4CHKJSYLPIL/graph.json","events_json":"https://pith.science/api/pith-number/63NRELHFFN5IYRD4CHKJSYLPIL/events.json","paper":"https://pith.science/paper/63NRELHF"},"agent_actions":{"view_html":"https://pith.science/pith/63NRELHFFN5IYRD4CHKJSYLPIL","download_json":"https://pith.science/pith/63NRELHFFN5IYRD4CHKJSYLPIL.json","view_paper":"https://pith.science/paper/63NRELHF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.13875&json=true","fetch_graph":"https://pith.science/api/pith-number/63NRELHFFN5IYRD4CHKJSYLPIL/graph.json","fetch_events":"https://pith.science/api/pith-number/63NRELHFFN5IYRD4CHKJSYLPIL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/63NRELHFFN5IYRD4CHKJSYLPIL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/63NRELHFFN5IYRD4CHKJSYLPIL/action/storage_attestation","attest_author":"https://pith.science/pith/63NRELHFFN5IYRD4CHKJSYLPIL/action/author_attestation","sign_citation":"https://pith.science/pith/63NRELHFFN5IYRD4CHKJSYLPIL/action/citation_signature","submit_replication":"https://pith.science/pith/63NRELHFFN5IYRD4CHKJSYLPIL/action/replication_record"}},"created_at":"2026-05-17T23:39:19.263000+00:00","updated_at":"2026-05-17T23:39:19.263000+00:00"}