English Calibration Erodes Lithuanian in Expert-Pruned Mixture-of-Experts Language Models
cs.CL, cs.LG
Submitted: 2026-11-05
Updated: 2026-11-06
Code: https://github.com/vllm-project/llm-compressor
Terminology
Sources
- Multilingual Routing in Mixture-of-Experts
- Mixture of Experts for Low-Resource LLMs
- Understanding Multilingualism in Mixture-of-Experts LLMs: Routing Mechanism, Expert Specialization, and Layerwise Steering
- Gemma 4 Technical Report
- GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models
- Half the Experts, All the Code: One-Shot Domain Pruning of Mixture-of-Experts LLMs for Coding
- REAM: Merging Improves Pruning of Experts in LLMs
- Leveraging Routing Dynamics in Mixture-of-Experts Models for Efficient Language Adaptation
- REAP the Experts: Why Pruning Prevails for One-Shot MoE compression
- The Uneven Impact of Post-Training Quantization in Machine Translation
- No Language Left Behind: Scaling Human-Centered Machine Translation
- Towards Robust Calibration Dataset for Domain-General Mixture of Experts Pruning
- Unveiling Super Experts in Mixture-of-Experts Large Language Models
- The Benchmark Illusion: Pruned LLMs Can Pass Multiple Choice but Fail to Answer
- Rethinking Calibration for High-Sparsity LLM Pruning: From Single Sources to Capability Coverage
- On the Utility and Factual Reliability of Pruned Mixture-of-Experts Models in the Biomedical Domain
- Qwen3 Technical Report
Related papers
- ALARM: Audio-Language Alignment for Reasoning Models
- ThinkSum: Probabilistic reasoning over sets using large language models
- Coherence boosting: When your pretrained language model is not paying enough attention
- From Facts to Insights: A Persona-Driven Dual Memory Framework and Dataset for Role-Playing Agents
- Improving General Role-Playing Agents via Psychology-Grounded Reasoning and Role-Aware Policy Optimization
- The Art of Asking: Multilingual Prompt Optimization for Synthetic Data