Router-norm changes induced by lightweight fine-tuning, for example router-only LoRA, provide a one-shot expert-pruning score that preserves MMLU-Pro accuracy far better than magnitude or random pruning on Mixtral-8x7B.
A provably effective method for pruning experts in fine-tuned sparse mixture-of-experts
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
method 1
citation-polarity summary
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1roles
method 1polarities
use method 1representative citing papers
citing papers explorer
-
Router Sensitivity Under Lightweight Fine-Tuning Identifies Prunable Experts in Mixture-of-Experts Models
Router-norm changes induced by lightweight fine-tuning, for example router-only LoRA, provide a one-shot expert-pruning score that preserves MMLU-Pro accuracy far better than magnitude or random pruning on Mixtral-8x7B.