{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:YI5XZ7QZ5YWAE7IZCUDQWARTMO","short_pith_number":"pith:YI5XZ7QZ","canonical_record":{"source":{"id":"2503.04647","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-06T17:33:01Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"9ceec1fd0af7b4a4625625c41ea003cf18bfeafb73a4f0d0acdafeb616055ac5","abstract_canon_sha256":"29bca3ccc1580aff54122926b4eede5a735d1b08b8e76b746f283e0160b3205a"},"schema_version":"1.0"},"canonical_sha256":"c23b7cfe19ee2c027d1915070b023363807edf2e02166d6e95fc6a4b072a48a6","source":{"kind":"arxiv","id":"2503.04647","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.04647","created_at":"2026-07-05T11:16:06Z"},{"alias_kind":"arxiv_version","alias_value":"2503.04647v2","created_at":"2026-07-05T11:16:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.04647","created_at":"2026-07-05T11:16:06Z"},{"alias_kind":"pith_short_12","alias_value":"YI5XZ7QZ5YWA","created_at":"2026-07-05T11:16:06Z"},{"alias_kind":"pith_short_16","alias_value":"YI5XZ7QZ5YWAE7IZ","created_at":"2026-07-05T11:16:06Z"},{"alias_kind":"pith_short_8","alias_value":"YI5XZ7QZ","created_at":"2026-07-05T11:16:06Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:YI5XZ7QZ5YWAE7IZCUDQWARTMO","target":"record","payload":{"canonical_record":{"source":{"id":"2503.04647","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-06T17:33:01Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"9ceec1fd0af7b4a4625625c41ea003cf18bfeafb73a4f0d0acdafeb616055ac5","abstract_canon_sha256":"29bca3ccc1580aff54122926b4eede5a735d1b08b8e76b746f283e0160b3205a"},"schema_version":"1.0"},"canonical_sha256":"c23b7cfe19ee2c027d1915070b023363807edf2e02166d6e95fc6a4b072a48a6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:16:06.819038Z","signature_b64":"Gz44ov43z6a0BFbJIS0WCp+A9JE+NTyRCUQO8/kFnTy6qug0cCYo6NtKhfVbxa5gMhJ6Pm2YGvdcRlcilARoDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c23b7cfe19ee2c027d1915070b023363807edf2e02166d6e95fc6a4b072a48a6","last_reissued_at":"2026-07-05T11:16:06.818427Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:16:06.818427Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.04647","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:16:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LcsUfBshd9G5y693WYhvd8LspnZFCkKYfeT4f9qICM8lA5PmZMpjOaTxxuofunfeVswRwuxqDQH3dp2y4AzuDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T22:30:10.330411Z"},"content_sha256":"1781a42398d9854f19a43ae2616092a09412cb7acfe1c6dd94939f6608161a10","schema_version":"1.0","event_id":"sha256:1781a42398d9854f19a43ae2616092a09412cb7acfe1c6dd94939f6608161a10"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:YI5XZ7QZ5YWAE7IZCUDQWARTMO","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Implicit Cross-Lingual Rewarding for Efficient Multilingual Preference Alignment","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chengqing Zong, Chen Wang, Jiajun Zhang, Junhong Wu, Wen Yang","submitted_at":"2025-03-06T17:33:01Z","abstract_excerpt":"Direct Preference Optimization (DPO) has become a prominent method for aligning Large Language Models (LLMs) with human preferences. While DPO has enabled significant progress in aligning English LLMs, multilingual preference alignment is hampered by data scarcity. To address this, we propose a novel approach that $\\textit{captures}$ learned preferences from well-aligned English models by implicit rewards and $\\textit{transfers}$ them to other languages through iterative training. Specifically, we derive an implicit reward model from the logits of an English DPO-aligned model and its correspon"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.04647","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/2503.04647/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:16:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kEHHS3snXvK9rZzt5Twe/Zk2gd6c9urKwuK+qgCyd6GeWjjHnc1ZqjFY40E8XNcdHka+JmN4OVGBTzEjBadiAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T22:30:10.331090Z"},"content_sha256":"8e4045152da7e42e022b01a2aa26d051d831fe5a5b9b618650ea32b3fbe5ad47","schema_version":"1.0","event_id":"sha256:8e4045152da7e42e022b01a2aa26d051d831fe5a5b9b618650ea32b3fbe5ad47"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/YI5XZ7QZ5YWAE7IZCUDQWARTMO/bundle.json","state_url":"https://pith.science/pith/YI5XZ7QZ5YWAE7IZCUDQWARTMO/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/YI5XZ7QZ5YWAE7IZCUDQWARTMO/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-09T22:30:10Z","links":{"resolver":"https://pith.science/pith/YI5XZ7QZ5YWAE7IZCUDQWARTMO","bundle":"https://pith.science/pith/YI5XZ7QZ5YWAE7IZCUDQWARTMO/bundle.json","state":"https://pith.science/pith/YI5XZ7QZ5YWAE7IZCUDQWARTMO/state.json","well_known_bundle":"https://pith.science/.well-known/pith/YI5XZ7QZ5YWAE7IZCUDQWARTMO/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:YI5XZ7QZ5YWAE7IZCUDQWARTMO","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"29bca3ccc1580aff54122926b4eede5a735d1b08b8e76b746f283e0160b3205a","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-06T17:33:01Z","title_canon_sha256":"9ceec1fd0af7b4a4625625c41ea003cf18bfeafb73a4f0d0acdafeb616055ac5"},"schema_version":"1.0","source":{"id":"2503.04647","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.04647","created_at":"2026-07-05T11:16:06Z"},{"alias_kind":"arxiv_version","alias_value":"2503.04647v2","created_at":"2026-07-05T11:16:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.04647","created_at":"2026-07-05T11:16:06Z"},{"alias_kind":"pith_short_12","alias_value":"YI5XZ7QZ5YWA","created_at":"2026-07-05T11:16:06Z"},{"alias_kind":"pith_short_16","alias_value":"YI5XZ7QZ5YWAE7IZ","created_at":"2026-07-05T11:16:06Z"},{"alias_kind":"pith_short_8","alias_value":"YI5XZ7QZ","created_at":"2026-07-05T11:16:06Z"}],"graph_snapshots":[{"event_id":"sha256:8e4045152da7e42e022b01a2aa26d051d831fe5a5b9b618650ea32b3fbe5ad47","target":"graph","created_at":"2026-07-05T11:16:06Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2503.04647/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Direct Preference Optimization (DPO) has become a prominent method for aligning Large Language Models (LLMs) with human preferences. While DPO has enabled significant progress in aligning English LLMs, multilingual preference alignment is hampered by data scarcity. To address this, we propose a novel approach that $\\textit{captures}$ learned preferences from well-aligned English models by implicit rewards and $\\textit{transfers}$ them to other languages through iterative training. Specifically, we derive an implicit reward model from the logits of an English DPO-aligned model and its correspon","authors_text":"Chengqing Zong, Chen Wang, Jiajun Zhang, Junhong Wu, Wen Yang","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-06T17:33:01Z","title":"Implicit Cross-Lingual Rewarding for Efficient Multilingual Preference Alignment"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.04647","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:1781a42398d9854f19a43ae2616092a09412cb7acfe1c6dd94939f6608161a10","target":"record","created_at":"2026-07-05T11:16:06Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"29bca3ccc1580aff54122926b4eede5a735d1b08b8e76b746f283e0160b3205a","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-06T17:33:01Z","title_canon_sha256":"9ceec1fd0af7b4a4625625c41ea003cf18bfeafb73a4f0d0acdafeb616055ac5"},"schema_version":"1.0","source":{"id":"2503.04647","kind":"arxiv","version":2}},"canonical_sha256":"c23b7cfe19ee2c027d1915070b023363807edf2e02166d6e95fc6a4b072a48a6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c23b7cfe19ee2c027d1915070b023363807edf2e02166d6e95fc6a4b072a48a6","first_computed_at":"2026-07-05T11:16:06.818427Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:16:06.818427Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Gz44ov43z6a0BFbJIS0WCp+A9JE+NTyRCUQO8/kFnTy6qug0cCYo6NtKhfVbxa5gMhJ6Pm2YGvdcRlcilARoDw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:16:06.819038Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.04647","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1781a42398d9854f19a43ae2616092a09412cb7acfe1c6dd94939f6608161a10","sha256:8e4045152da7e42e022b01a2aa26d051d831fe5a5b9b618650ea32b3fbe5ad47"],"state_sha256":"340aa7e39f5182a6e68ba765457140ed35cb817071346f505b5a410d6f06e82d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"R1Z+ClhZSmB7k/P5Osyq3Lw7kk01TJxh3+v7vXMdRJJJoxL+5SMkI3f6SOJ3gtq4ngp1wyZ1+S8khiytb/EqDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T22:30:10.337105Z","bundle_sha256":"d710731460a7cf74a300a16a94b1896828bda34c7a447b75877302e6cb8efb5f"}}