Hot eventLive
DRMoET：分布鲁棒MoE训练方法
1 reports1 sources3 hr ago updated
Get the story
AI overview
2026年10月7日，研究者在 arXiv（Machine Learning Theory，一手报道）提出 Distributionally Robust MoE Training（DRMoET）。该方法把逐层专家视为内生鲁棒组，优化高损失路由结果，而非仅均衡流量。目前公开信息仅涉及方法思路，未见实验数据、代码或同行评议进展。
Generated from reports · updated 3 hr ago
LatestOct 7
2026-10-07，DRMoET在arXiv提出，以分布鲁棒方式优化高损失路由结果。Timeline
Follow the coverage from different angles.
Oct 7, 2026
- arXiv · Machine Learning Theory分布鲁棒 Mixture-of-Experts 训练
研究者提出 Distributionally Robust MoE Training（DRMoET），把逐层专家视为内生鲁棒组，优化高损失路由结果而非仅均衡流量。
Heat trend
Current heat 9·Comparable peak 10（Oct 7）·Comparable change over 24 hours –
The trend compares only the same participants observed continuously; its range may be smaller than the current heat count. Move or click on the chart to inspect hourly heat; use the left and right arrow keys to switch.