연구
Detecting a Route Flip Is Easier Than Knowing Whether to Fix It: Causal Route-Mediated Damage in Quantized Mixture-of-Experts
arXiv:2608.11212v1 Announce Type: new Abstract: Topk MixtureofExperts MoE routing is discontinuous, so a deploymentmotivated numerical disturbance simulated 4bit KVcache quantization read by a protected BF16 gate pushes tokens across decision boundaries and flips which experts fire.
이 콘텐츠는 ArXiv AI 원본 기사의 요약입니다. 전문은 원본 사이트에서 확인해주세요.
원문 기사 보기 →