Strengthening Target-Language Features: SAE-Based Steering for Multilingual Inference

• URL: http://arxiv.org/abs/2608.04904v2
• Date: Thu, 06 Aug 2026 04:47:27 GMT
• Title: Strengthening Target-Language Features: SAE-Based Steering for Multilingual Inference
• Abstract summary: We propose an inference-time multilingual steering method that uses pretrained sparse autoencoders to identify and strengthen target-language-related features.<n>Experiments with Gemma-3-12B-it show average accuracy improvements of 10.9 percentage points on XCOPA, 5.3 points on XNLI, and 1.9 points on MGSM.
• Score: 10.43186930809146
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Multilingual large language models exhibit substantial performance differences across languages, while existing adaptation methods often require parameter updates and considerable multilingual training data. We propose an inference-time multilingual steering method that uses pretrained sparse autoencoders to identify and strengthen target-language-related features. Using multilingual parallel sentences, we compare SAE activations across languages and select a small number of layer-specific features associated with each target language. These features are decoded into steering signals and injected into the model's hidden states without additional training. Experiments with Gemma-3-12B-it show average accuracy improvements of 10.9 percentage points on XCOPA, 5.3 points on XNLI, and 1.9 points on MGSM.

Related papers

• Multilingual Steering by Design: Multilingual Sparse Autoencoders and Principled Layer Selection [16.857692737043198]
  We show that training SAEs on multilingual data consistently strengthens cross-lingual representations.<n>We introduce an empha priori steering layer-selection rule based on the intersection of multilingual alignment and language separability.<n>Our results show that multilingual SAEs combined with intersection-selected layers stabilize the trade-off between language identification accuracy and generation quality.
  arXiv  Detail & Related papers  (2026-05-21T21:00:32Z)
• LangGPS: Language Separability Guided Data Pre-Selection for Joint Multilingual Instruction Tuning [49.22807995935406]
  Joint multilingual instruction tuning is a widely adopted approach to improve the multilingual instruction-following ability and downstream performance of large language models (LLMs)<n>Existing selection methods, often based on features like text quality, diversity, or task relevance, typically overlook the intrinsic linguistic structure of multilingual data.<n>We propose LangGPS, a lightweight two-stage pre-selection framework guided by language separability which quantifies how well samples in different languages can be distinguished in the model's representation space.
  arXiv  Detail & Related papers  (2025-11-13T12:02:32Z)
• Causal Language Control in Multilingual Transformers via Sparse Feature Steering [7.754609745940422]
  We investigate whether sparse autoencoder features can be leveraged to steer the generated language of multilingual language models.<n>We achieve controlled language shifts with up to 90% success, as measured by FastText language classification.<n>Our analysis reveals that language steering is most effective in mid-to-late transformer layers.
  arXiv  Detail & Related papers  (2025-07-17T06:49:16Z)
• Enhancing Multilingual ASR for Unseen Languages via Language Embedding Modeling [50.62091603179394]
  Whisper, one of the most advanced ASR models, handles 99 languages effectively.<n>However, Whisper struggles with unseen languages, those not included in its pre-training.<n>We propose methods that exploit these relationships to enhance ASR performance on unseen languages.
  arXiv  Detail & Related papers  (2024-12-21T04:05:43Z)
• Adapting the adapters for code-switching in multilingual ASR [10.316724084739892]
  Large pre-trained multilingual speech models have shown potential in scaling Automatic Speech Recognition to many low-resource languages. Some of these models employ language adapters in their formulation, which helps to improve monolingual performance. This formulation restricts the usability of these models on code-switched speech, where two languages are mixed together in the same utterance. We propose ways to effectively fine-tune such models on code-switched speech, by assimilating information from both language adapters at each language adaptation point in the network.
  arXiv  Detail & Related papers  (2023-10-11T12:15:24Z)
• Soft Language Clustering for Multilingual Model Pre-training [57.18058739931463]
  We propose XLM-P, which contextually retrieves prompts as flexible guidance for encoding instances conditionally. Our XLM-P enables (1) lightweight modeling of language-invariant and language-specific knowledge across languages, and (2) easy integration with other multilingual pre-training methods.
  arXiv  Detail & Related papers  (2023-06-13T08:08:08Z)
• xGQA: Cross-Lingual Visual Question Answering [100.35229218735938]
  xGQA is a new multilingual evaluation benchmark for the visual question answering task. We extend the established English GQA dataset to 7 typologically diverse languages. We propose new adapter-based approaches to adapt multimodal transformer-based models to become multilingual.
  arXiv  Detail & Related papers  (2021-09-13T15:58:21Z)
• Are Multilingual Models Effective in Code-Switching? [57.78477547424949]
  We study the effectiveness of multilingual language models to understand their capability and adaptability to the mixed-language setting. Our findings suggest that pre-trained multilingual models do not necessarily guarantee high-quality representations on code-switching.
  arXiv  Detail & Related papers  (2021-03-24T16:20:02Z)
• Learning to Scale Multilingual Representations for Vision-Language Tasks [51.27839182889422]
  The effectiveness of SMALR is demonstrated with ten diverse languages, over twice the number supported in vision-language tasks to date. We evaluate on multilingual image-sentence retrieval and outperform prior work by 3-4% with less than 1/5th the training parameters compared to other word embedding methods.
  arXiv  Detail & Related papers  (2020-04-09T01:03:44Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.