# Knowledge Localization in Mixture-of-Experts LLMs Using Cross-Lingual Inconsistency

arXiv: https://arxiv.org/abs/2603.17102 · HTML: https://mlsanity.com/paper/2603.17102/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Placed in MoE large language models, one of 12 directions in Mixture-of-experts.
- Builds on 38 of its 48 references that are in the map.
- First posted on arXiv 17 Mar 2026.

## Placed in

- Mixture-of-experts / [MoE large language models](https://mlsanity.com/moe/moe-large-language-models/)

## Builds on (most cited first, 30 of 38)

- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer](https://mlsanity.com/paper/1701.06538/) (2017)
- [Axiomatic Attribution for Deep Networks](https://mlsanity.com/paper/1703.01365/) (2017)
- [Locating and Editing Factual Associations in GPT](https://mlsanity.com/paper/2202.05262/) (2022)
- [Sparse Autoencoders Find Highly Interpretable Features in Language Models](https://mlsanity.com/paper/2309.08600/) (2023)
- [Mass-Editing Memory in a Transformer](https://mlsanity.com/paper/2210.07229/) (2022)
- [Dissecting Recall of Factual Associations in Auto-Regressive Language Models](https://mlsanity.com/paper/2304.14767/) (2023)
- [Layer by Layer: Uncovering Hidden Representations in Language Models](https://mlsanity.com/paper/2502.02013/) (2025)
- [Does Localization Inform Editing? Surprising Differences in Causality-Based Localization vs. Knowledge Editing in Language Models](https://mlsanity.com/paper/2301.04213/) (2023)
- [OpenMoE: An Early Effort on Open Mixture-of-Experts Language Models](https://mlsanity.com/paper/2402.01739/) (2024)
- [PMET: Precise Model Editing in a Transformer](https://mlsanity.com/paper/2308.08742/) (2023)
- [Cross-Lingual Consistency of Factual Knowledge in Multilingual Language Models](https://mlsanity.com/paper/2310.10378/) (2023)
- [INCLUDE: Evaluating Multilingual Language Understanding with Regional Knowledge](https://mlsanity.com/paper/2411.19799/) (2024)
- [On the Multilingual Ability of Decoder-based Pre-trained Language Models: Finding and Controlling Language-Specific Neurons](https://mlsanity.com/paper/2404.02431/) (2024)
- [The Semantic Hub Hypothesis: Language Models Share Semantic Representations Across Languages and Modalities](https://mlsanity.com/paper/2411.04986/) (2024)
- [A Rigorous Study of Integrated Gradients Method and Extensions to Internal Neuron Attributions](https://mlsanity.com/paper/2202.11912/) (2022)
- [Demons in the Detail: On Implementing Load Balancing Loss for Training Specialized Mixture-of-Expert Models](https://mlsanity.com/paper/2501.11873/) (2025)
- [A Closer Look into Mixture-of-Experts in Large Language Models](https://mlsanity.com/paper/2406.18219/) (2024)
- [Separating Tongue from Thought: Activation Patching Reveals Language-Agnostic Concept Representations in Transformers](https://mlsanity.com/paper/2411.08745/) (2024)
- [Do Localization Methods Actually Localize Memorized Data in LLMs? A Tale of Two Benchmarks](https://mlsanity.com/paper/2311.09060/) (2023)
- [ECLeKTic: a Novel Challenge Set for Evaluation of Cross-Lingual Knowledge Transfer](https://mlsanity.com/paper/2502.21228/) (2025)
- [Multilingual Routing in Mixture-of-Experts](https://mlsanity.com/paper/2510.04694/) (2025)
- [MultiLoKo: a multilingual local knowledge benchmark for LLMs spanning 31 languages](https://mlsanity.com/paper/2504.10356/) (2025)
- [Steering MoE LLMs via Expert (De)Activation](https://mlsanity.com/paper/2509.09660/) (2025)
- [Language-Specific Latent Process Hinders Cross-Lingual Performance](https://mlsanity.com/paper/2505.13141/) (2025)
- [Paths Not Taken: Understanding and Mending the Multilingual Factual Recall Pipeline](https://mlsanity.com/paper/2505.20546/) (2025)
- [The Emergence of Abstract Thought in Large Language Models Beyond Any Language](https://mlsanity.com/paper/2506.09890/) (2025)
- [Probing Semantic Routing in Large Mixture-of-Expert Models](https://mlsanity.com/paper/2502.10928/) (2025)
- [Tracing Multilingual Factual Knowledge Acquisition in Pretraining](https://mlsanity.com/paper/2505.14824/) (2025)
- [Does Knowledge Localization Hold True? Surprising Differences Between Entity and Relation Perspectives in Language Models](https://mlsanity.com/paper/2409.00617/) (2024)

## Built on by (most cited first, 6 of 6)

- [Extracting Small Translation Specialists from LLMs by Aggressively Pruning Experts](https://mlsanity.com/paper/2605.28042/) (2026)
- [Cross-Lingual Steering for Figurative Language Generation](https://mlsanity.com/paper/2605.30443/) (2026)
- [A theoretical model for task routing in mixture-of-expert transformers](https://mlsanity.com/paper/2606.14398/) (2026)
- [LMs as Task-Specific Knowledge Bases: An Interpretability Analysis](https://mlsanity.com/paper/2606.27237/) (2026)
- [Multilinguality in Hybrid Attention LLMs](https://mlsanity.com/paper/2609.35378/) (2026)
- [Cross-Lingual Alignment for Decoder-Only Models using MoE Routers](https://mlsanity.com/paper/2610.01921/) (2026)

## Abstract (arXiv)

Modern LLMs continue to exhibit significant variance in behavior across languages, such as being able to recall factual information in some languages but not others. While typically studied as a problem to be mitigated, in this work, we propose leveraging this cross-lingual inconsistency as a tool for interpretability in mixture-of-experts (MoE) LLMs. Our knowledge localization framework contrasts routing for sets of languages where the model correctly recalls information from languages where it fails. This allows us to isolate model components that play a functional role in answering about a piece of knowledge. Our method proceeds in two stages: (1) querying the model with difficult factual questions across a diverse set of languages to generate "success" and "failure" activation buckets and then (2) applying a statistical contrastive analysis to the MoE router logits to identify experts important for knowledge. To validate the necessity of this small number of experts for answering a knowledge question, we deactivate them and re-ask the question. We find that despite only deactivating about 20 out of 6000 experts, the model no longer answers correctly in over 40% of cases. Generally, this method provides a realistic and scalable knowledge localization approach to address increasingly complex LLMs.