{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/when-is-bert-multilingual-isolating-crucial","title":"When is BERT Multilingual? Isolating Crucial Ingredients for Cross-lingual Transfer","arxiv_id":"2110.14782","date":"2021-10-27","proceeding":"NAACL 2022 7","authors":["Ameet Deshpande","Partha Talukdar","Karthik Narasimhan"],"abstract":"While recent work on multilingual language models has demonstrated their capacity for cross-lingual zero-shot transfer on downstream tasks, there is a lack of consensus in the community as to what shared properties between languages enable such transfer. Analyses involving pairs of natural languages are often inconclusive and contradictory since languages simultaneously differ in many linguistic aspects. In this paper, we perform a large-scale empirical study to isolate the effects of various linguistic properties by measuring zero-shot transfer between four diverse natural languages and their counterparts constructed by modifying aspects such as the script, word order, and syntax. Among other things, our experiments show that the absence of sub-word overlap significantly affects zero-shot transfer when languages differ in their word order, and there is a strong correlation between transfer performance and word embedding alignment between languages (e.g., R=0.94 on the task of NLI). Our results call for focus in multilingual models on explicitly improving word embedding alignment between languages rather than relying on its implicit emergence.","url_abs":"https://arxiv.org/abs/2110.14782v3","url_pdf":"https://arxiv.org/pdf/2110.14782v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"when-is-bert-multilingual-isolating-crucial","repo_url":"https://github.com/princeton-nlp/MultilingualAnalysis","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"jax","reach":null},{"paper_slug":"when-is-bert-multilingual-isolating-crucial","repo_url":"https://github.com/princeton-nlp/align-mlm","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"jax","reach":null}],"tasks":[{"task_slug":"cross-lingual-transfer","task_name":"Cross-Lingual Transfer"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=2110.14782","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2110.14782"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/princeton-nlp/MultilingualAnalysis","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/princeton-nlp/align-mlm","reach":null}],"summary":{"ran_draft_wrong":5,"ran_fixture":1},"by_repo_kind":{"official":{"samples":5,"ran":5,"repositories":1},"listed":{"samples":1,"ran":1,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"9a2d9f4b00fd2506","entry":"add_hyperparams_get_command","repo":"princeton-nlp/MultilingualAnalysis","repo_kind":"official","path":"run_experiments.py","file_url":"https://github.com/princeton-nlp/MultilingualAnalysis/blob/HEAD/run_experiments.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9a2d9f4b00fd2506"}},{"code_sha256_prefix":"a2d02277323f02b3","entry":"construct_command","repo":"princeton-nlp/MultilingualAnalysis","repo_kind":"official","path":"run_experiments.py","file_url":"https://github.com/princeton-nlp/MultilingualAnalysis/blob/HEAD/run_experiments.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a2d02277323f02b3"}},{"code_sha256_prefix":"48576e423c05dd22","entry":"get_mlm_command","repo":"princeton-nlp/MultilingualAnalysis","repo_kind":"official","path":"run_experiments.py","file_url":"https://github.com/princeton-nlp/MultilingualAnalysis/blob/HEAD/run_experiments.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"48576e423c05dd22"}},{"code_sha256_prefix":"52dc9cb4e8c95988","entry":"get_normalized_embeddings","repo":"princeton-nlp/align-mlm","repo_kind":"listed","path":"analysis/alignedmlm_paper_analysis/compute_alignment_with_lang_ids.py","file_url":"https://github.com/princeton-nlp/align-mlm/blob/HEAD/analysis/alignedmlm_paper_analysis/compute_alignment_with_lang_ids.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"52dc9cb4e8c95988"}},{"code_sha256_prefix":"86d49a2d45aea79c","entry":"get_orthogonal_mapping","repo":"princeton-nlp/MultilingualAnalysis","repo_kind":"official","path":"analysis/learn_orthogonal_mapping_one_one.py","file_url":"https://github.com/princeton-nlp/MultilingualAnalysis/blob/HEAD/analysis/learn_orthogonal_mapping_one_one.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"86d49a2d45aea79c"}},{"code_sha256_prefix":"1cea4fd97708b2a8","entry":"get_train_validation_indices","repo":"princeton-nlp/MultilingualAnalysis","repo_kind":"official","path":"analysis/learn_orthogonal_mapping_one_one.py","file_url":"https://github.com/princeton-nlp/MultilingualAnalysis/blob/HEAD/analysis/learn_orthogonal_mapping_one_one.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"1cea4fd97708b2a8"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}