avatar
ABLingss
UESTC undergraduate researcher in compiler, MLSys, CV, and LLM engineering. First-author papers @ NeurIPS / MICCAI 2026. Open-source contributor to HeartMuLa (4.2k+ ★).

# ABLingss

Undergraduate researcher in Machine Learning Systems, AI Compilers, Computer Vision, and Audio Generation.

I care about building systems that work — efficient inference, compiler optimizations, and reproducible research.

# Education

University of Electronic Science and Technology of China (UESTC) — Sept 2024 – Present

# Research Interests

  • ML Systems & Efficient Inference: CUDA optimization, FlashAttention, INT8 quantization, KV-cache sparsification, streaming inference
  • AI Compilers: LLVM IR passes (loop/SSA/scalar optimizations), MLIR, TVM AutoScheduler
  • Audio Generation: AR music foundation models, latent diffusion, codec design for long-form generation
  • Computer Vision: Medical image analysis, unsupervised anatomical representation learning

# Technical Strengths

Area Technologies
ML Systems & Inference CUDA (shared memory, thread blocks), CUDA Graph, Triton, FlashAttention, vLLM, ONNX Runtime, INT8/PTQ, KV-cache optimization, NCCL, Ray
Compilers & Systems LLVM IR passes (loop/SSA/scalar), MLIR, TVM AutoScheduler, Clang LibTooling, Flex/Bison
Programming C/C++ (CMake), Python, PyTorch, Transformers

# Contact