Deep Learning
A³: Adaptive Audio Agent Framework for Collaborative Edge Computing
Embedded audio agents need a nearby edge server for speed, but a server-dependent agent returns nothing when the link fails. We present …
Zexin Li
,
Xueyi Zhang
,
Tao Ren
,
Yiming Chen
PDF
Cite
DOI
RLᴿᵀ: On-Device Reinforcement Learning Runtime
On-device reinforcement learning runs a periodic actor and a learner on one embedded GPU shared with other workloads. Because the …
Zexin Li
,
Xinwei Gao
,
Arvind Easwaran
PDF
Cite
DOI
ProCerDCL: Query-Efficient Black-Box Probabilistic Verification of Deep Learning Compiler Optimization
We present ProCerDCL, a query-efficient framework for black-box probabilistic verification. Given a fixed, sampleable input …
Zexin Li
,
Weibo Zhang
,
Linyi Li
,
Arvind Easwaran
,
Simin Chen
PDF
Cite
DOI
Boosting Transferable Adversarial Attacks against Deep Reinforcement Learning
Most adversarial attacks on deep reinforcement learning (DRL) assume white-box access to the victim policy, which rarely holds in …
Zexin Li
,
Ruili Yao
,
Yiming Zeng
,
Xiaoxue Gao
PDF
Cite
DOI
Real-Time Continual Learning on Embedded GPUs
Online continual learning (OCL) is an enabling technology for safety-critical autonomous embedded systems whose deployment environments …
Zexin Li
,
Dongjoo Seo
,
Nikil Dutt
,
Jun Sheng
,
Cong Liu
PDF
Cite
VLMSysTrojan: Stealthy System-Aware Backdoor Attacks Against Vision-Language Models
Vision-Language Models (VLMs) are increasingly deployed in real-world applications. To ensure efficient inference, these deployments …
Yezheng Cheng
,
Zexin Li
,
Jiaqi Wu
,
Zhihong Zhang
,
Simin Chen
PDF
Cite
Towards Deployable On-Device Machine Learning Systems
Safety-critical cyber-physical systems (CPS), such as autonomous vehicles and robots, increasingly rely on deep neural networks. While …
Zexin Li
PDF
Cite
Source Document
TreeDiff: AST-Guided Code Generation with Diffusion LLMs
Code generation is increasingly critical for real-world applications. Still, diffusion-based large language models continue to struggle …
Yiming Zeng
,
Jinghan Cao
,
Zexin Li
,
Yiming Chen
,
Tao Ren
,
Zhuochun Li
,
Dawei Xiang
,
Xidong Wu
,
Shangqian Gao
,
Tingting Yu
PDF
Cite
DOI
HyperEdit: Unlocking Instruction-based Text Editing in LLMs via Hypernetworks
Instruction-based text editing is increasingly critical for real-world applications such as code editors (e.g., Cursor), but Large …
Yiming Zeng
,
Jinghan Cao
,
Zexin Li
,
Wanhao Yu
,
Zhankai Ye
,
Dawei Xiang
,
Ting Hua
,
Xin Liu
,
Shangqian Gao
,
Tingting Yu
PDF
Cite
NaturalSloth: Revisiting Denial-of-Service Attacks on Large Language Models
LLM serving is limited by provider-side resources: longer generations consume more GPU time, increase latency, and reduce throughput in …
Yiming Chen
,
Zexin Li
,
Xianghu Yue
,
Robby T. Tan
,
Haizhou Li
PDF
Cite
DOI
»
Cite
×