mlsanityData to 7 Oct 2026

Cross-Lingual Prompt Steerability: Towards Accurate and Robust LLM Behavior across Languages

arXiv 2512.02841 ↗First posted 2 Dec 2025cs.CL cs.AI cs.HCcited by 3

Not on the map. Why?

Abstract (arXiv)

System prompts provide a lightweight yet powerful mechanism for conditioning large language models (LLMs) at inference time. While prior work has focused on English-only settings, real-world deployments benefit from having a single prompt to operate reliably across languages. This paper presents a comprehensive study of how different system prompts steer models toward accurate and robust cross-lingual behavior. We propose a unified four-dimensional evaluation framework to assess system prompts in multilingual environments. Through large-scale experiments on five languages, three LLMs, and three benchmarks, we uncover that certain prompt components, such as CoT, emotion, and scenario, correlate with robust multilingual behavior. We develop a prompt optimization framework for multilingual settings and show it can automatically discover prompts that improve all metrics by 5-10%. Finally, we analyze over 10 million reasoning units and find that more performant system prompts induce more structured and consistent reasoning patterns, while reducing unnecessary language-switching. Together, we highlight system prompt optimization as a scalable path to accurate and robust multilingual LLM behavior.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 26 of its 50 references that are in the map. First posted on arXiv 2 Dec 2025.

Builds on

26 of 50 references.

References · 26

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  2. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  3. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,767
  4. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,203
  5. 2019Unsupervised Cross-lingual Representation Learning at Scalecited by 3,907
  6. 2023Let's Verify Step by Stepcited by 3,598
  7. 2021TruthfulQA: Measuring How Models Mimic Human Falsehoodscited by 2,919
  8. 2024MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmarkcited by 1,758
  9. 2022Measuring and Narrowing the Compositionality Gap in Language Modelscited by 1,054
  10. 2016Bag of Tricks for Efficient Text Classificationcited by 869
  11. 2024Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inferencecited by 527
  12. 2025Cognitive Behaviors that Enable Self-Improving Reasoners, or, Four Habits of Highly Effective STaRscited by 373
14 more
  1. 2022Reasoning with Language Model Prompting: A Surveycited by 251
  2. 2021Reframing Instructional Prompts to GPTk's Languagecited by 176
  3. 2023Not All Languages Are Created Equal in LLMs: Improving Multilingual Capability by Cross-Lingual-Thought Promptingcited by 163
  4. 2023Large Language Models Understand and Can be Enhanced by Emotional Stimulicited by 143
  5. 2023Rephrase and Respond: Let Large Language Models Ask Better Questions for Themselvescited by 123
  6. 2023Large Language Models as Analogical Reasonerscited by 107
  7. 2024Can Large Language Models Reason and Plan?cited by 82
  8. 2024Prompting Techniques for Reducing Social Bias in LLMs through System 1 and System 2 Cognitive Processescited by 38
  9. 2025Inference-Time Computations for LLM Reasoning and Planning: A Benchmark and Insightscited by 30
  10. 2024SPRIG: Improving Large Language Model Performance by System Prompt Optimizationcited by 28
  11. 2025Language Matters: How Do Multilingual Input and Reasoning Paths Affect Large Reasoning Models?cited by 20
  12. 2025Are Rules Meant to be Broken? Understanding Multilingual Moral Reasoning as a Computational Pipeline with UniMoralcited by 14
  13. 2025GAAPO: Genetic Algorithmic Applied to Prompt Optimizationcited by 10
  14. 2024Native vs Non-Native Language Prompting: A Comparative Analysiscited by 9

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Untangling Input Language from Reasoning Language: A Diagnostic Framework for Cross-Lingual Moral Alignment in LLMscited by 1
  2. 2026Imperative Interference: Social Register Shapes Instruction Topology in Large Language Modelscited by 0
  3. 2026Re-Centering Humans in LLM Personalizationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025The Art of Asking: Multilingual Prompt Optimization for Synthetic Data
  2. 2024GReaTer: Gradients over Reasoning Makes Smaller Language Models Strong Prompt Optimizers
  3. 2025Self-Supervised Prompt Optimization
  4. 2025PromptBridge: Cross-Model Prompt Transfer for Large Language Models
  5. 2025How and Where to Translate? The Impact of Translation Strategies in Cross-lingual LLM Prompting
  6. 2024The Language Barrier: Dissecting Safety Challenges of LLMs in Multilingual Contexts
  7. 2024A Systematic Survey of Prompt Engineering in Large Language Models: Techniques and Applications
  8. 2025PromptCoT 2.0: Scaling Prompt Synthesis for Large Language Model Reasoning