{"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/understanding-and-mitigating-language","title":"Understanding and Mitigating Language Confusion in LLMs","arxiv_id":"2406.20052","date":"2024-06-28","proceeding":null,"authors":["Kelly Marchisio","Wei-Yin Ko","Alexandre Bérard","Théo Dehaze","Sebastian Ruder"],"abstract":"We investigate a surprising limitation of LLMs: their inability to consistently generate text in a user's desired language. We create the Language Confusion Benchmark (LCB) to evaluate such failures, covering 15 typologically diverse languages with existing and newly-created English and multilingual prompts. We evaluate a range of LLMs on monolingual and cross-lingual generation reflecting practical use cases, finding that Llama Instruct and Mistral models exhibit high degrees of language confusion and even the strongest models fail to consistently respond in the correct language. We observe that base and English-centric instruct models are more prone to language confusion, which is aggravated by complex prompts and high sampling temperatures. We find that language confusion can be partially mitigated via few-shot prompting, multilingual SFT and preference tuning. We release our language confusion benchmark, which serves as a first layer of efficient, scalable multilingual evaluation at https://github.com/for-ai/language-confusion.","url_abs":"https://arxiv.org/abs/2406.20052v2","url_pdf":"https://arxiv.org/pdf/2406.20052v2.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":"understanding-and-mitigating-language","repo_url":"https://github.com/for-ai/language-confusion","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[],"methods":[{"method_slug":"base","method_name":"BASE"},{"method_slug":"llama","method_name":"LLaMA"},{"method_slug":"sft","method_name":"SFT"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2406.20052","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2406.20052"}},"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/for-ai/language-confusion","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"ran":3},"by_repo_kind":{"official":{"samples":3,"ran":3,"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":"a996a61a010f6934","entry":"normalize","repo":"for-ai/language-confusion","repo_kind":"official","path":"compute_metrics.py","file_url":"https://github.com/for-ai/language-confusion/blob/HEAD/compute_metrics.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"a996a61a010f6934"}},{"code_sha256_prefix":"3f7e1abc53ea4130","entry":"tokenize","repo":"for-ai/language-confusion","repo_kind":"official","path":"compute_metrics.py","file_url":"https://github.com/for-ai/language-confusion/blob/HEAD/compute_metrics.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"3f7e1abc53ea4130"}},{"code_sha256_prefix":"8ad8ac8b774652af","entry":"turns_to_prompt","repo":"for-ai/language-confusion","repo_kind":"official","path":"few_shot_prompting.py","file_url":"https://github.com/for-ai/language-confusion/blob/HEAD/few_shot_prompting.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"8ad8ac8b774652af"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}