| Challenge: | Large language models suffer from language confusion, a phenomenon in which responses are partially or entirely generated in unintended languages. |
| Approach: | They propose a supervised fine-tuning methodology which optimizes the likelihood of correct tokens without explicitly penalizing undesired outputs such as cross-lingual mixing. |
| Outcome: | The proposed model suppresses language-confused generation while maintaining strong language consistency even under high decoding temperatures while preserving general QA performance. |
Similar Papers
Understanding and Mitigating Language Confusion in LLMs (2024.emnlp-main)
Copied to clipboard
| Challenge: | Llama Instruct and Mistral models exhibit high degrees of language confusion and even the strongest models fail to consistently respond in the correct language. |
| Approach: | They develop a language confusion benchmark to evaluate LLMs' inability to consistently generate text in a user’s desired language. |
| Outcome: | The proposed model fails to consistently respond in the correct language, despite being prone to high temperatures and complex prompts. |
TLPO: Token-Level Policy Optimization for Mitigating Language Confusion in Large Language Models (2026.acl-long)
Copied to clipboard
| Challenge: | Prior mitigation approaches that optimize entire responses operate at the level of entire responses and can lead to unintended degradation of general model capabilities. |
| Approach: | They propose a fine-tuning framework to mitigate erroneous outputs by localizing and updating the policy at a granular level. |
| Outcome: | The proposed framework outperforms baselines on multiple multilingual LLMs across diverse languages while preserving task accuracy. |
Mitigating Language Confusion through Inference-time Intervention (2025.coling-main)
Copied to clipboard
| Challenge: | Existing methods to address the problem of language confusion are incontext learning and supervised fine-tuning (SFT) however, they consume context window space and require extensive data collection. |
| Approach: | They propose a language-sensitive intervention that detects and assesses language confusion without additional complex mechanisms. |
| Outcome: | The proposed method detects language confusion and assesses content quality without additional complex mechanisms. |
The Language Barrier: Dissecting Safety Challenges of LLMs in Multilingual Contexts (2024.findings-acl)
Copied to clipboard
Lingfeng Shen, Weiting Tan, Sihao Chen, Yunmo Chen, Jingyu Zhang, Haoran Xu, Boyuan Zheng, Philipp Koehn, Daniel Khashabi
| Challenge: | Recent studies show that malicious prompt instructions could solicit objectionable content from LLMs. |
| Approach: | They compare how state-of-the-art LLMs respond to malicious prompts in different languages . they find that LLM's generate unsafe responses more often when a prompt is written in a lower-resource language . |
| Outcome: | The proposed model can generate unsafe responses more often when a malicious prompt is written in a lower-resource language, and less irrelevant responses when written in lower-source languages. |
Investigating the Multilingual Calibration Effects of Language Model Instruction Tuning (2026.eacl-short)
Copied to clipboard
Jerry Huang, Peng Lu, Qiuhao Zeng, Yusuke Iwasawa, Yutaka Matsuo, Sarath Chandar, Edison Marrese-Taylor, Irene Li
| Challenge: | despite advances in foundation model research, the relationship between large language models and their calibration remains an open area of research. |
| Approach: | They examine a gap in the calibration of large language models within multilingual settings to better understand how data scarcity can potentially lead to different calibration effects. |
| Outcome: | The proposed calibration gap is found in two multilingual benchmarks over 29 and 42 languages. |
Evaluating Robustness of Large Language Models Against Multilingual Typographical Errors (2026.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) are increasingly deployed in multilingual, real-world applications where user inputs introduce typographical errors. |
| Approach: | They propose a multilingual typo generation algorithm that simulates human-like errors based on language-specific keyboard layouts and typing behavior. |
| Outcome: | The proposed model can generate the correct answer ("500") under typos in English, German, and Russian. |
VEEF-Multi-LLM: Effective Vocabulary Expansion and Parameter Efficient Finetuning Towards Multilingual Large Language Models (2025.coling-main)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have a significant disadvantage for low-resource languages . VEEF-Multi-LLM-8B excels in multilingual instruction-following tasks . |
| Approach: | They propose a low-resource multilingual large language model that expands the vocabulary for multilingual support. |
| Outcome: | The proposed model outperforms existing models in multilingual instruction-following tasks, but lags behind English-centric models in some tasks. |
Fine-tuning Large Language Models with Limited Data: A Survey and Practical Guide (2026.tacl-1)
Copied to clipboard
| Challenge: | Pre-trained language models provide strong foundations, but effective adaptation under data scarcity requires efficient and efficient fine-tuning techniques. |
| Approach: | They propose to review parameter-efficient fine-tuning techniques that lower training and deployment costs and domain and cross-lingual adaptation methods for both encoder and decoder models. |
| Outcome: | The proposed techniques lower training and deployment costs, domain and cross-lingual adaptation methods, and model specialization strategies. |
Towards Understanding the Fragility of Multilingual LLMs against Fine-Tuning Attacks (2025.findings-naacl)
Copied to clipboard
| Challenge: | Recent advances in Large Language Models have sparked concerns about their safety. |
| Approach: | They propose a method to identify safety-related information in the model parameter space . they propose to use a few adversarially chosen examples to fine-tune LLMs . |
| Outcome: | The proposed method can break safety alignment in multilingual LLMs using a few examples . it also shows that the proposed method jailbreaks LLM models adapted to new languages . |
ReCoVeR the Target Language: Language Steering without Sacrificing Task Performance (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Large Language Models exhibit more language confusion as they become multilingual . authors propose a lightweight approach for reducing language confusion based on language-specific steering vectors . |
| Approach: | They propose a lightweight approach to reduce language confusion by using language-specific steering vectors. |
| Outcome: | The proposed approach reduces language confusion in large language models . it leverages language-specific steering vectors for effective LLM steering . |