# MetaSICL: Globalizing Auditory LLMs for Underserved Speakers and Languages via Meta Speech In-Context Learning

arXiv: https://arxiv.org/abs/2601.18904 · HTML: https://mlsanity.com/paper/2601.18904/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 24 references that are in the map.
- First posted on arXiv 26 Jan 2026.

## Builds on (most cited first, 21 of 21)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Common Voice: A Massively-Multilingual Speech Corpus](https://mlsanity.com/paper/1912.06670/) (2019)
- [Qwen2.5-Omni Technical Report](https://mlsanity.com/paper/2503.20215/) (2025)
- [Language Is Not All You Need: Aligning Perception with Language Models](https://mlsanity.com/paper/2302.14045/) (2023)
- [FLEURS: Few-shot Learning Evaluation of Universal Representations of Speech](https://mlsanity.com/paper/2205.12446/) (2022)
- [MetaICL: Learning to Learn In Context](https://mlsanity.com/paper/2110.15943/) (2021)
- [Audio Flamingo: A Novel Audio Language Model with Few-Shot Learning and Dialogue Abilities](https://mlsanity.com/paper/2402.01831/) (2024)
- [MMSU: A Massive Multi-task Spoken Language Understanding and Reasoning Benchmark](https://mlsanity.com/paper/2506.04779/) (2025)
- [SALM: Speech-augmented Language Model with In-context Learning for Speech Recognition and Translation](https://mlsanity.com/paper/2310.09424/) (2023)
- [Can Whisper perform speech-based in-context learning?](https://mlsanity.com/paper/2309.07081/) (2023)
- [COSMIC: Data Efficient Instruction-tuning For Speech In-Context Learning](https://mlsanity.com/paper/2311.02248/) (2023)
- [My Science Tutor (MyST) -- A Large Corpus of Children's Conversational Speech](https://mlsanity.com/paper/2309.13347/) (2023)
- [UniAudio 1.5: Large Language Model-driven Audio Codec is A Few-shot Audio Task Learner](https://mlsanity.com/paper/2406.10056/) (2024)
- [The Interspeech 2025 Speech Accessibility Project Challenge](https://mlsanity.com/paper/2507.22047/) (2025)
- [Bayesian Example Selection Improves In-Context Learning for Speech, Text, and Visual Modalities](https://mlsanity.com/paper/2404.14716/) (2024)
- [SMILE: Speech Meta In-Context Learning for Low-Resource Language Automatic Speech Recognition](https://mlsanity.com/paper/2409.10429/) (2024)
- [M2R-Whisper: Multi-stage and Multi-scale Retrieval Augmentation for Enhancing Whisper](https://mlsanity.com/paper/2409.11889/) (2024)
- [TICL+: A Case Study On Speech In-Context Learning for Children's Speech Recognition](https://mlsanity.com/paper/2512.18263/) (2025)
- [Few-shot Personalization via In-Context Learning for Speech Emotion Recognition based on Speech-Language Model](https://mlsanity.com/paper/2509.08344/) (2025)
- [TICL: Text-Embedding KNN For Speech In-Context Learning Unlocks Speech Recognition Abilities of Large Multimodal Models](https://mlsanity.com/paper/2509.13395/) (2025)

## Built on by (most cited first, 1 of 1)

- [FSA-GRPO: Teaching Auditory LLMs to Use Few-Shot Demonstrations](https://mlsanity.com/paper/2606.02615/) (2026)

## Abstract (arXiv)

Generative AI for speech and audio is increasingly expected to serve users across languages, cultures, and communities, yet current auditory Large Language Models (LLMs) are still largely trained and evaluated on high-resource data. Globalizing such systems requires handling low-resource settings, where the target speakers, languages, or tasks are poorly represented in training data. In these regimes, collecting enough labeled in-domain data is often impractical, and the small corpora available may still under-represent the test distribution, making direct fine-tuning brittle under domain shift. In-Context Learning (ICL) offers an alternative: instead of updating model parameters for every underserved community, an auditory LLM can adapt at inference time by conditioning on a few local demonstrations. However, vanilla speech ICL remains limited because most auditory LLMs are not explicitly trained to use such demonstrations effectively. We address this gap with Meta Speech In-Context Learning (MetaSICL), a post-training recipe that strengthens an auditory LLM's in-context adaptation ability using only abundant high-resource speech data. Although MetaSICL never trains on the target low-resource domains, it improves performance across two backbones on children's ASR, audio understanding/reasoning, and speech translation and ASR in directions and languages unseen in post-training. We further study the case where some in-domain data is available, using low-resource language ASR as a case study, since recognition for underserved languages is central to globalizing generative AI. Here, using MetaSICL as a warmup for in-domain reinforcement learning yields the strongest results, outperforming direct fine-tuning across five typologically diverse languages. Overall, MetaSICL offers a practical route toward globalizing auditory LLMs by building inference-time adaptation into the model.