SlimWise: Decoupling Expert Pruning Across Prefill and Decode for Efficient MoE Serving
Paper • 2609.34117 • Published • 12 •
4
Gunho Park
gunho1123
AI & ML interests
None yet
Recent Activity
commentedon a paper about 5 hours ago
SlimWise: Decoupling Expert Pruning Across Prefill and Decode for Efficient MoE Serving authored a paper about 21 hours ago
No Token Left Behind: Reliable KV Cache Compression via Importance-Aware
Mixed Precision Quantization authored a paper about 21 hours ago
CodeGEMM: A Codebook-Centric Approach to Efficient GEMM in Quantized LLMsOrganizations
None yet
commented a
paper about 5 hours ago
authored 2
papers about 21 hours ago
authored 5
papers about 22 hours ago
Affine-Scaled Attention: Towards Flexible and Stable Transformer Attention
Paper • 2602.23057 • Published
Training-free Dropout Sampling for Semantic Token Acceptance in Speculative Decoding
Paper • 2603.03333 • Published
LUT-GEMM: Quantized Matrix Multiplication based on LUTs for Efficient Inference in Large-Scale Generative Language Models
Paper • 2206.09557 • Published
An Investigation of FP8 Across Accelerators for LLM Inference
Paper • 2502.01070 • Published • 3
Faster Inference of LLMs using FP8 on the Intel Gaudi
Paper • 2503.09975 • Published • 1
upvoted a paper about 22 hours ago
submitted a
paper to Daily Papers about 22 hours ago
authored a
paper about 22 hours ago
updated 5
models 8 months ago
gunho1123/Llama-3.1-8B-Instruct-Codegemm-m2v8g128
2B • Updated • 15
gunho1123/phi-4-anybcq2to4-g128
Text Generation • 4B • Updated • 13
gunho1123/gemma-2-9b-anybcq2to4-g128
Text Generation • 3B • Updated • 14
gunho1123/Llama-3.1-8B-anybcq2to4-g128
3B • Updated • 20
gunho1123/Qwen3-32B-anybcq2to4-g128
9B • Updated • 12
updated a
collection 8 months ago
published 2
models 8 months ago