Making Every Verified Token Count: Adaptive Verification for MoE Speculative Decoding
EVICT is proposed, a training-free, hyperparameter-free, and lossless adaptive verification method for MoE speculative decoding that achieves up to 2.35x speedup over autoregressive decoding and an average 1.21x speedup over the state-of-the-art baseline EAGLE-3, while significantly reducing unnecessary expert activati...