# How a Bilingual LM Becomes Bilingual: Tracing Internal Representations with Sparse Autoencoders

arXiv: https://arxiv.org/abs/2503.06394 · HTML: https://mlsanity.com/paper/2503.06394/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Placed in SAE features and their applications, one of 13 directions in Mechanistic interpretability.
- Builds on 11 of its 21 references that are in the map.
- First posted on arXiv 9 Mar 2025.

## Placed in

- Mechanistic interpretability / [SAE features and their applications](https://mlsanity.com/mechinterp/sae-features-and-their-applications/)

## Builds on (most cited first, 11 of 11)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Sparse Autoencoders Find Highly Interpretable Features in Language Models](https://mlsanity.com/paper/2309.08600/) (2023)
- [k-Sparse Autoencoders](https://mlsanity.com/paper/1312.5663/) (2013)
- [On the Multilingual Ability of Decoder-based Pre-trained Language Models: Finding and Controlling Language-Specific Neurons](https://mlsanity.com/paper/2404.02431/) (2024)
- [Sharing Matters: Analysing Neurons Across Languages and Tasks in LLMs](https://mlsanity.com/paper/2406.09265/) (2024)
- [LLM-jp: A Cross-organizational Project for the Research and Development of Fully Open Japanese LLMs](https://mlsanity.com/paper/2407.03963/) (2024)
- [Towards Universality: Studying Mechanistic Similarity Across Language Model Architectures](https://mlsanity.com/paper/2410.06672/) (2024)
- [Converging to a Lingua Franca: Evolution of Linguistic Regions and Semantics Alignment in Multilingual Large Language Models](https://mlsanity.com/paper/2410.11718/) (2024)
- [Mechanistic Permutability: Match Features Across Layers](https://mlsanity.com/paper/2410.07656/) (2024)
- [Tracking the Feature Dynamics in LLM Training: A Mechanistic Study](https://mlsanity.com/paper/2412.17626/) (2024)
- [Evolution of SAE Features Across Layers in LLMs](https://mlsanity.com/paper/2410.08869/) (2024)

## Built on by (most cited first, 5 of 5)

- [Visual Exploration of Feature Relationships in Sparse Autoencoders with Curated Concepts](https://mlsanity.com/paper/2511.06048/) (2025)
- [When Do Concepts Become Functionally Sufficient During Language-Model Training?](https://mlsanity.com/paper/2608.15323/) (2026)
- [Features have life history. And we should care](https://mlsanity.com/paper/2605.18789/) (2026)
- [When Safety Speaks a Language: A Mechanistic Analysis of Safety-Language Identity Entanglement in LLMs](https://mlsanity.com/paper/2608.29936/) (2026)
- [Beetle: A Bilingual Model Suite for Modelling Second-Language Processing](https://mlsanity.com/paper/2609.22633/) (2026)

## Abstract (arXiv)

This study explores how bilingual language models develop complex internal representations. We employ sparse autoencoders to analyze internal representations of bilingual language models with a focus on the effects of training steps, layers, and model sizes. Our analysis shows that language models first learn languages separately, and then gradually form bilingual alignments, particularly in the mid layers. We also found that this bilingual tendency is stronger in larger models. Building on these findings, we demonstrate the critical role of bilingual representations in model performance by employing a novel method that integrates decomposed representations from a fully trained model into a mid-training model. Our results provide insights into how language models acquire bilingual capabilities.