Seoul National University VLSI Lab

university

https://vlsi.snu.ac.kr/

Activity Feed Request to join this org

AI & ML interests

Efficient AI

Recent Activity

jiwonsong authored a paper 7 days ago

Joint Subspace Sparse Attention for Efficient Long-Context Decoding

jiwonsong submitted a paper 7 days ago

Joint Subspace Sparse Attention for Efficient Long-Context Decoding

Hyeongju97 authored a paper 23 days ago

WaveFront Decoding: Parallelized Self-Speculative Decoding for Looped Language Models

View all activity

Papers

Joint Subspace Sparse Attention for Efficient Long-Context Decoding

PReCache: Efficient KV Cache Sharing for Multi-LoRA Agents via Low-Rank Precomputation and Neutral Reconstruction

View all Papers

SNU-VLSI 's papers 11

Submitted by

Jiwon Song

Joint Subspace Sparse Attention for Efficient Long-Context Decoding

SNU-VLSI

Seoul National University VLSI Lab

Submitted by

Hyesung Jeon

PReCache: Efficient KV Cache Sharing for Multi-LoRA Agents via Low-Rank Precomputation and Neutral Reconstruction

SNU-VLSI

Seoul National University VLSI Lab

Submitted by

Hyesung Jeon

KVCMAS: Efficient KV cache Correction for Shared Context in Multi-Agent Systems

SNU-VLSI

Seoul National University VLSI Lab

Submitted by

Hyeongju Ha

WaveFront Decoding: Parallelized Self-Speculative Decoding for Looped Language Models

SNU-VLSI

Seoul National University VLSI Lab

Submitted by

Jiwon Song

CompactAttention: Accelerating Chunked Prefill with Block-Union KV Selection

SNU-VLSI

Seoul National University VLSI Lab

Submitted by

Jiwon Song

RelayGen: Intra-Generation Model Switching for Efficient Reasoning

SNU-VLSI

Seoul National University VLSI Lab

Submitted by

Jiwon Song

LiteStage: Latency-aware Layer Skipping for Multi-stage Reasoning

SNU-VLSI

Seoul National University VLSI Lab

2

Retrospective Sparse Attention for Efficient Long-Context Generation

SNU-VLSI

Seoul National University VLSI Lab

Submitted by

Jiwon Song

Reasoning Path Compression: Compressing Generation Trajectories for Efficient LLM Reasoning

SNU-VLSI

Seoul National University VLSI Lab

SLEB: Streamlining LLMs through Redundancy Verification and Elimination of Transformer Blocks

SNU-VLSI

Seoul National University VLSI Lab

Squeezing Large-Scale Diffusion Models for Mobile

SNU-VLSI

Seoul National University VLSI Lab