On the Utility and Factual Reliability of Pruned Mixture-of-Experts Models in the Biomedical Domain

Mixture-of-Experts (MoE) models offer inference speedups via selective activation but impose substantial memory requirements because the whole network must remain loaded. Structured expert pruning is a practical approach for reducing deployment costs in resource-constrained settings. However, prior studies primarily evaluate benchmark utility, leaving the effect of pruning on factual reliability underexplored, particularly in high-stakes domains such as biomedicine. In this paper, we investigate how ___domain-specific expert pruning affects both utility and reliability. We assess four MoE models, six pruning methods, and multiple pruning ratios across generation and classification tasks under in-___domain (biomedical) and cross-___domain settings. Results reveal that moderate pruning preserves in-___domain utility without immediate reliability decline, although hallucination risks increase at extreme pruning ratios. When shifting to the general ___domain, both utility and reliability degrade rapidly. These findings indicate that safe compression depends heavily on the task and ___domain. Evaluating pruned MoE models solely on utility is inadequate for high-stakes deployment without reliability assessment.