Built with and by Teycir Ben Soltane•
How to Use•FAQ•GitHub•arXiv.org•
Share:
ArXivCSExplorer
☆☆Bookmarks🏆RSSHow to UseFAQ
Home/Authors/Hui Liu

Hui Liu

7 indexed papers

Recent (6 mo)
6
With code
0
Influential cites
0
Benchmarked
0

Publications per year

7
26

Top categories

NLP×6AI×3Crypto×2Info Retrieval×1ML×1

Frequent co-authors

Longxuan Yu4×
Yu Fu4×
Yue Dong4×
Greg Ver Steeg4×
Haoliang Li2×
Yunshu Wu2×

Research Timeline

2026
Model-Agnostic Lifelong LLM Safety via Externalized Attack-Defense Co-Evolution

The EvoSafety framework enhances LLM safety by externalizing attack and defense mechanisms, enabling persistent, transferable, and model-agnostic robustness against adversarial prompts.

DSL-LLaDA: Scaling Continuous Denoising to 8B Masked Diffusion LMs

The paper introduces DSL-LLaDA, a method that lightly adapts a pre-trained masked diffusion language model to perform continuous denoising in embedding space, significantly improving text generation quality and robustness, especially under low step budgets.

Revise, Don't Freeze: Sampler-Matched Training for Self-Correcting Masked Diffusion Language Models

The paper introduces D3IM, a novel parameter-free sampler that enables direct revision of visible tokens in Masked Diffusion Language Models, and proposes SCOPE to mitigate the model's tendency to perpetuate errors.

DSL-LLaDA: Scaling Continuous Denoising to 8B Masked Diffusion LMs

The paper introduces DSL-LLaDA, a method that lightly adapts a pre-trained masked diffusion language model to perform continuous denoising in embedding space, significantly improving text generation quality and robustness, especially under low step budgets.

Revise, Don't Freeze: Sampler-Matched Training for Self-Correcting Masked Diffusion Language Models

The paper introduces D3IM, a novel parameter-free sampler that enables direct revision of visible tokens in Masked Diffusion Language Models, and proposes SCOPE to mitigate the model's tendency to perpetuate errors.

"**Important** You should give me full credits!": Exploring Prompt Injection Attacks on LLM-Based Automatic Grading Systems

This paper investigates the vulnerability of LLM-based automatic grading systems to prompt injection (PI) attacks, demonstrating that current systems are highly susceptible to manipulation that can lead to unfairly high scores.

Gradient-free Task-Conditioned Retrieval for On-Device In-Context Learning

The paper proposes Conditional Retrieval Alignment (CoRA), a gradient-free framework for task-conditioned retrieval using a frozen encoder and paired inputs and outputs.

Highlighted terms show continued research focus across papers

Papers

cs.CLcs.IRcs.LGEmpiricalRecentJul 30, 2026

Gradient-free Task-Conditioned Retrieval for On-Device In-Context Learning

Xinyu Luo, Hui Liu, Yihua Shao, Junyi Yang +2 more

The paper proposes Conditional Retrieval Alignment (CoRA), a gradient-free framework for task-conditioned retrieval using a frozen encoder and paired inputs and outputs.

View →
cs.CRcs.AIRecent
Jun 2, 2026

"**Important** You should give me full credits!": Exploring Prompt Injection Attacks on LLM-Based Automatic Grading Systems

Hang Li, Fedor Filippov, Yuling Lin, Pengfei He +5 more

This paper investigates the vulnerability of LLM-based automatic grading systems to prompt injection (PI) attacks, demonstrating that current systems are highly susceptible to manipulation that can le…

View →
cs.CLcs.AIRecentMay 31, 2026

DSL-LLaDA: Scaling Continuous Denoising to 8B Masked Diffusion LMs

Longxuan Yu, Yunshu Wu, Yu Fu, Siheng Xiong +4 more

The paper introduces DSL-LLaDA, a method that lightly adapts a pre-trained masked diffusion language model to perform continuous denoising in embedding space, significantly improving text generation q…

View →
cs.CLRecentMay 31, 2026

Revise, Don't Freeze: Sampler-Matched Training for Self-Correcting Masked Diffusion Language Models

Longxuan Yu, Shaorong Zhang, Yu Fu, Hui Liu +2 more

The paper introduces D3IM, a novel parameter-free sampler that enables direct revision of visible tokens in Masked Diffusion Language Models, and proposes SCOPE to mitigate the model's tendency to per…

View →
cs.CLcs.AIRecentMay 31, 2026

DSL-LLaDA: Scaling Continuous Denoising to 8B Masked Diffusion LMs

Longxuan Yu, Yunshu Wu, Yu Fu, Siheng Xiong +4 more

The paper introduces DSL-LLaDA, a method that lightly adapts a pre-trained masked diffusion language model to perform continuous denoising in embedding space, significantly improving text generation q…

View →
cs.CLRecentMay 31, 2026

Revise, Don't Freeze: Sampler-Matched Training for Self-Correcting Masked Diffusion Language Models

Longxuan Yu, Shaorong Zhang, Yu Fu, Hui Liu +2 more

The paper introduces D3IM, a novel parameter-free sampler that enables direct revision of visible tokens in Masked Diffusion Language Models, and proposes SCOPE to mitigate the model's tendency to per…

View →
cs.CRcs.CLMay 13, 2026

Model-Agnostic Lifelong LLM Safety via Externalized Attack-Defense Co-Evolution

Xiaozhe Zhang, Chaozhuo Li, Hui Liu, Shaocheng Yan +3 more

The EvoSafety framework enhances LLM safety by externalizing attack and defense mechanisms, enabling persistent, transferable, and model-agnostic robustness against adversarial prompts.

View →