publications
publications by categories in reversed chronological order. generated by jekyll-scholar.
2026
- Under Review
HERALD: High Throughput Block Diffusion LLM Inference via CPU-GPU Cooperative KV Cache RetrievalUnder Review, 2026 - Under ReviewTriage: Skip, Skim, or Refine Attention Tiles in Video Diffusion TransformersUnder Review, 2026
- NeurIPS
MAGE: All-[MASK] Block Already Knows Where to Look in Block Diffusion LLMNeurIPS 2026
ICML 2026 Adaptive Foundation Model Workshop (Oral, Best Paper Award (1/218 = 0.46%))
2025
- ICCAD
STAR: Improving Lifetime and Performance of High-Capacity Modern SSDs Using State-Aware RandomizerICCAD 2025 (Oral)U.S. Patent Application with SK hynix (No. 19/569899; filed March 17, 2026) - CAL
AiDE: Attention-FFN Disaggregated Execution for Cost-Effective LLM Decoding on CXL-PNMIEEE CAL, 2025