OLA: Output Language Alignment in Code-Switched LLM Interactions

Code-switching, alternating between languages within a conversation, is natural for multilingual users, yet poses fundamental challenges for large language models (LLMs). When a user code-switches in their prompt to an LLM, they typically do not specify the expected language of the LLM response, and thus LLMs must infer the output language from contextual and pragmatic cues. We find that current LLMs systematically fail to align with this expectation, responding in undesired languages even when cues are clear to humans. We introduce OLA, a benchmark to evaluate LLMs' Output Language Alignment in codeswitched interactions. OLA focuses on Korean-English code-switching and spans simple intrasentential mixing to instruction-content mismatches. Even frontier models frequently misinterpret implicit language expectation, exhibiting a bias toward non-English responses. We further show this bias generalizes beyond Korean to Chinese and Indonesian pairs. Models also show instability through mid-response switching and language intrusions. Chain-of-Thought prompting fails to resolve these errors, indicating weak pragmatic reasoning about output language. However, Code-Switching Aware DPO with minimal data (∼1K examples) substantially reduces misalignment, suggesting these failures stem from insufficient alignment rather than fundamental limitations. Our results highlight the need to align multilingual LLMs with users' implicit expectations in real-world code-switched interactions. 1 * Equal contribution. 1 OLA is available at https://github.com/juhyunohh/ OLA replies to be in the language of the original message → Response language: English 안녕하세요. 본 운동(Main Workout) 전에 워밍업 세트(Warm-up Sets)를 수행하는 것은 부상을 방지하고 운동 효과를 [...] Translation: Hello. Performing warm-up sets before the main workout is very important for preventing injuries and maximizing [...] )