# Improving the Downstream Performance of Mixture-of-Experts Transformers via Weak Vanilla Transformers

arXiv: https://arxiv.org/abs/2403.01994 · HTML: https://mlsanity.com/paper/2403.01994/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Expert routing strategies, one of 12 directions in Mixture-of-experts.
- Builds on 8 of its 26 references that are in the map.
- First posted on arXiv 4 Mar 2024.

## Placed in

- Mixture-of-experts / [Expert routing strategies](https://mlsanity.com/moe/expert-routing-strategies/)

## Builds on (most cited first, 8 of 8)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [Scaling Laws for Neural Language Models](https://mlsanity.com/paper/2001.08361/) (2020)
- [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://mlsanity.com/paper/1910.01108/) (2019)
- [The Pile: An 800GB Dataset of Diverse Text for Language Modeling](https://mlsanity.com/paper/2101.00027/) (2020)
- [FastMoE: A Fast Mixture-of-Expert Training System](https://mlsanity.com/paper/2103.13262/) (2021)
- [Mixture-of-Experts Meets Instruction Tuning:A Winning Combination for Large Language Models](https://mlsanity.com/paper/2305.14705/) (2023)

## Abstract (arXiv)

Recently, Mixture of Experts (MoE) Transformers have garnered increasing attention due to their advantages in model capacity and computational efficiency. However, studies have indicated that MoE Transformers underperform vanilla Transformers in many downstream tasks, significantly diminishing the practical value of MoE models. To explain this issue, we propose that the pre-training performance and transfer capability of a model are joint determinants of its downstream task performance. MoE models, in comparison to vanilla models, have poorer transfer capability, leading to their subpar performance in downstream tasks. To address this issue, we introduce the concept of transfer capability distillation, positing that although vanilla models have weaker performance, they are effective teachers of transfer capability. The MoE models guided by vanilla models can achieve both strong pre-training performance and transfer capability, ultimately enhancing their performance in downstream tasks. We design a specific distillation method and conduct experiments on the BERT architecture. Experimental results show a significant improvement in downstream performance of MoE models, and many further evidences also strongly support the concept of transfer capability distillation. Finally, we attempt to interpret transfer capability distillation and provide some insights from the perspective of model feature.