Lu Jingyu · 卢景瑜

ANU  ·  UCD '24  ·  BJUT '24

Lu Jingyu
About

Postgraduate researcher at Australian National University, advised by Prof. Shoaib Akram. My research focuses on computer architecture and systems — currently on prefetcher optimisation and memory interference modelling. Previously at University College Dublin and Beijing University of Technology (Beijing-Dublin International College), where I graduated with First Class Honours.

I have also built independent projects in parallel computing and CUDA development, and contribute to LLM safety research through Nvidia Garak.

Research Interests
Computer Architecture Systems Programming OS Internals HPC Distributed Systems GPU Computing LLM Safety
Education
Australian National University
Postgraduate · 2024 –
University College Dublin
First Class Honours · 2024
Beijing University of Technology
Graduate · 2024
Publications
SciSafeEval: A Comprehensive Benchmark for Safety Alignment of Large Language Models in Scientific Tasks
T Li, J Lu, C Chu, T Zeng, Y Zheng, M Li, H Huang, B Wu, Z Liu, K Ma, et al.
AAAI AICS 2024 Oral
SciSafeEval figure
Synthetic Poisoning Attacks: The Impact of Poisoned MRI Image on U-Net Brain Tumor Segmentation
T Li, T Zeng, Y Zheng, C Zhang, J Lu, H Huang, C Chu, F Yin, Z Yang
AAPM 2025  ·  ICLR Workshop 2025 Oral Poster
Synthetic Poisoning Attacks figure
Are All Prompt Components Value-Neutral? Understanding the Heterogeneous Adversarial Robustness of Dissected Prompt in Large Language Models
Y Zheng, T Li, H Huang, T Zeng, J Lu, C Chu, Y Huang, Z Jiang, Q Xiong, Y Ge, M Li
EACL 2025 Oral
Prompt Robustness figure
Towards Emotion-Aware Healthcare: A Comprehensive Review of Multimodal Emotion Recognition Technologies in Medical Practice
J Lu, Y Zheng, C Chu, T Zeng, Y Zhao, T Li
ICDT 2024 Oral
Multimodal Emotion Recognition figure
Projects
(Some academic projects are no longer publicly available due to access and academic integrity considerations, further details can be provided upon request via email)

🖥️  Systems Programming

xv6
xv6-os-projects 2024
Kernel Engineering · MIT xv6

Kernel-level extensions to MIT's xv6: system calls, scheduling algorithms, virtual memory management, and file system features — all implemented from scratch.

xv6
CSAPP-memory-allocators-and-garbage-collectors 2024
Memory Systems · CSAPP

Heap allocator with explicit free lists and boundary-tag coalescing, paired with a mark-and-sweep garbage collector. Optimised to balance throughput and fragmentation ratio.

train control
CSAPP-concurrent-train-traffic-control 2024
Concurrent Systems · CSAPP

Concurrent train scheduler with deadlock prevention via resource ordering. Fine-grained mutex strategy minimises lock contention across critical sections using POSIX threads.

⚡  Parallel & High-Performance Computing

CUDA solver
shared-memory-gpu-advection-solver 2024
GPU Computing · CUDA

GPU solver for advection equations using CUDA shared memory tiling. Warp-aligned data staging cuts global memory bandwidth and maximises on-chip reuse.

MPI solver
distributed-advection-solver-mpi 2024
Distributed Computing · MPI

Multi-node advection solver with halo exchange. Non-blocking MPI communication pipelines computation and data exchange across distributed processes.

🔐  LLM Safety

garak
nvidia garak 2024
LLM Safety · Red-teaming

Contributing to an open-source LLM vulnerability scanner — probing language models for jailbreaks, hallucination patterns, and safety boundary failures.