{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:QGYRHMFWMLYQOMPDKPPF6MVZ7N","short_pith_number":"pith:QGYRHMFW","schema_version":"1.0","canonical_sha256":"81b113b0b662f10731e353de5f32b9fb66d1b10a77216ba4f4dd021606cc16db","source":{"kind":"arxiv","id":"2304.09151","version":1},"attestation_state":"computed","paper":{"title":"UniMax: Fairer and more Effective Language Sampling for Large-Scale Multilingual Pretraining","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Adam Roberts, Hyung Won Chung, Noah Constant, Orhan Firat, Sharan Narang, Xavier Garcia, Yi Tay","submitted_at":"2023-04-18T17:45:50Z","abstract_excerpt":"Pretrained multilingual large language models have typically used heuristic temperature-based sampling to balance between different languages. However previous work has not systematically evaluated the efficacy of different pretraining language distributions across model scales. In this paper, we propose a new sampling method, UniMax, that delivers more uniform coverage of head languages while mitigating overfitting on tail languages by explicitly capping the number of repeats over each language's corpus. We perform an extensive series of ablations testing a range of sampling strategies on a s"},"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":"2304.09151","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-04-18T17:45:50Z","cross_cats_sorted":[],"title_canon_sha256":"291662e6585d81856dcdef8c8476eabb574805364481299715eee3a8bd691ac2","abstract_canon_sha256":"398ad3e5ee063a8d44cec92a8816fa6c2293ed314047283e934eb3f34aae03a8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:02:17.674168Z","signature_b64":"YZGcZFrc/v3AfCMWozVbIpDCaQFraCbtbPh/ef3rLx2IPMfYNBe4oOEcJAN5hrDLSrhtTwjyyusgmPZ/Eza+BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"81b113b0b662f10731e353de5f32b9fb66d1b10a77216ba4f4dd021606cc16db","last_reissued_at":"2026-07-05T06:02:17.673553Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:02:17.673553Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"UniMax: Fairer and more Effective Language Sampling for Large-Scale Multilingual Pretraining","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Adam Roberts, Hyung Won Chung, Noah Constant, Orhan Firat, Sharan Narang, Xavier Garcia, Yi Tay","submitted_at":"2023-04-18T17:45:50Z","abstract_excerpt":"Pretrained multilingual large language models have typically used heuristic temperature-based sampling to balance between different languages. However previous work has not systematically evaluated the efficacy of different pretraining language distributions across model scales. In this paper, we propose a new sampling method, UniMax, that delivers more uniform coverage of head languages while mitigating overfitting on tail languages by explicitly capping the number of repeats over each language's corpus. We perform an extensive series of ablations testing a range of sampling strategies on a s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2304.09151","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/2304.09151/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":"2304.09151","created_at":"2026-07-05T06:02:17.673638+00:00"},{"alias_kind":"arxiv_version","alias_value":"2304.09151v1","created_at":"2026-07-05T06:02:17.673638+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2304.09151","created_at":"2026-07-05T06:02:17.673638+00:00"},{"alias_kind":"pith_short_12","alias_value":"QGYRHMFWMLYQ","created_at":"2026-07-05T06:02:17.673638+00:00"},{"alias_kind":"pith_short_16","alias_value":"QGYRHMFWMLYQOMPD","created_at":"2026-07-05T06:02:17.673638+00:00"},{"alias_kind":"pith_short_8","alias_value":"QGYRHMFW","created_at":"2026-07-05T06:02:17.673638+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":19,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08763","citing_title":"OpenCoF: Learning to Reason Through Video Generation","ref_index":5,"is_internal_anchor":true},{"citing_arxiv_id":"2606.26058","citing_title":"DomainShuttle: Freeform Open Domain Subject-driven Text-to-video Generation","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2606.17730","citing_title":"ActWorld: From Explorable to Interactive World Model via Action-Aware Memory","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2606.13676","citing_title":"Modality Forcing for Scalable Spatial Generation","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30514","citing_title":"3D Scene-Adaptive Trajectory-Controllable Human Image Animation with Camera Movement","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05290","citing_title":"Do Models Share Safety Representations? Cross-Model Steering for Safe Visual Generation","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30514","citing_title":"3D Scene-Adaptive Trajectory-Controllable Human Image Animation with Camera Movement","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2503.20314","citing_title":"Wan: Open and Advanced Large-Scale Video Generative Models","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17488","citing_title":"Omni-Customizer: End-to-End MultiModal Customization for Joint Audio-Video Generation","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2510.01186","citing_title":"ASTRA: Let Arbitrary Subjects Transform in Video Editing","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2512.15840","citing_title":"Large Video Planner Enables Generalizable Robot Control","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2512.20340","citing_title":"The devil is in the details: Enhancing Video Virtual Try-On via Keyframe-Driven Details Injection","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2512.21714","citing_title":"AstraNav-World: World Model for Foresight Control and Consistency","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2602.07775","citing_title":"Rolling Sink: Bridging Limited-Horizon Training and Open-Ended Testing in Autoregressive Video Diffusion","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2604.16380","citing_title":"Data Mixing for Large Language Models Pretraining: A Survey and Outlook","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2504.13074","citing_title":"SkyReels-V2: Infinite-length Film Generative Model","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06859","citing_title":"Knowledge Transfer Scaling Laws for 3D Medical Imaging","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08121","citing_title":"Uni-ViGU: Towards Unified Video Generation and Understanding via A Diffusion-Based Video Generator","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18648","citing_title":"DanceCrafter: Fine-Grained Text-Driven Controllable Dance Generation via Choreographic Syntax","ref_index":8,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QGYRHMFWMLYQOMPDKPPF6MVZ7N","json":"https://pith.science/pith/QGYRHMFWMLYQOMPDKPPF6MVZ7N.json","graph_json":"https://pith.science/api/pith-number/QGYRHMFWMLYQOMPDKPPF6MVZ7N/graph.json","events_json":"https://pith.science/api/pith-number/QGYRHMFWMLYQOMPDKPPF6MVZ7N/events.json","paper":"https://pith.science/paper/QGYRHMFW"},"agent_actions":{"view_html":"https://pith.science/pith/QGYRHMFWMLYQOMPDKPPF6MVZ7N","download_json":"https://pith.science/pith/QGYRHMFWMLYQOMPDKPPF6MVZ7N.json","view_paper":"https://pith.science/paper/QGYRHMFW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2304.09151&json=true","fetch_graph":"https://pith.science/api/pith-number/QGYRHMFWMLYQOMPDKPPF6MVZ7N/graph.json","fetch_events":"https://pith.science/api/pith-number/QGYRHMFWMLYQOMPDKPPF6MVZ7N/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QGYRHMFWMLYQOMPDKPPF6MVZ7N/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QGYRHMFWMLYQOMPDKPPF6MVZ7N/action/storage_attestation","attest_author":"https://pith.science/pith/QGYRHMFWMLYQOMPDKPPF6MVZ7N/action/author_attestation","sign_citation":"https://pith.science/pith/QGYRHMFWMLYQOMPDKPPF6MVZ7N/action/citation_signature","submit_replication":"https://pith.science/pith/QGYRHMFWMLYQOMPDKPPF6MVZ7N/action/replication_record"}},"created_at":"2026-07-05T06:02:17.673638+00:00","updated_at":"2026-07-05T06:02:17.673638+00:00"}