PPL
Hongyang Zhang
Papers (6)
EAGLE-2: Faster Inference of Language Models with Dynamic Draft TreesEAGLE-3: Scaling up Inference Acceleration of Large Language Models via Training-Time TestEAGLE: Speculative Sampling Requires Rethinking Feature UncertaintyD$^3$-Subsidy: Online and Sequential Driver Subsidy Decision-Making for Large-Scale Ride-Hailing MarketMPerS: Dynamic MLLM MixExperts Perception-Guided Remote Sensing Scene SegmentationAttention Drift: What Autoregressive Speculative Decoding Models Learn