Skip to content

2 papers indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Stabilizing MoE Reinforcement Learning by Aligning Training and Inference Routers

Rollout Routing Replay (R3), a method that records routing distributions from the inference engine and replays them during training significantly reduces training-inference policy KL divergence and mitigates extreme discrepancies without compromising training speed, is proposed.

Wenhan Ma, Hai-Lin Zhang, Liang Zhao et al. · 90 citations · ⚡17

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.