Summary

First-year Ph.D. student in ECE at Cornell University working on efficient and scalable systems for machine learning, with an emphasis on distributed training and inference.

Education

• Ph.D. in Electrical and Computer Engineering
  Cornell University
• B.S. in Applied Physics (EECS Department)

  2026-07
  Peking University

  GPA: 3.75/4.0 (Top 18%)

Skills

Publications

• CREATE: Cross-Layer Resilience Characterization and Optimization for Efficient yet Reliable Embodied AI Systems

  2026
  ASPLOS 2026

  Characterizes the reliability of LLM-based embodied AI systems at the application, system, and circuit levels, and proposes error detection and correction techniques for efficient yet reliable embodied AI. (Co-first author.)

  View Publication
• Faster-MoA: Low-Latency Tree-Structured MoA Serving with Early Exit and Agent-Aware Prefill-Decode Overlap

  2026
  DAC 2026

  Accelerates Mixture-of-Agents serving with a tree-structured aggregation architecture, semantic-similarity-based early exit, and agent-aware prefill-decode overlap. (Co-first author.)

  View Publication
• DySL-VLA: Efficient Vision-Language-Action Model Inference via Dynamic-Static Layer-Skipping for Robot Manipulation

  2026
  DAC 2026

  Adaptive layer-skipping for VLA model inference based on task requirements and motion significance, cutting trainable parameters by 85.7× versus full fine-tuning.

  View Publication