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

Lei Ma

4 indexed papers

Recent (6 mo)
4
With code
0
Influential cites
0
Benchmarked
0

Publications per year

4
26

Top categories

AI×4Crypto×2NLP×1

Frequent co-authors

Yi Liu2×
Gelei Deng2×
Yuekang Li2×
Leo Yu Zhang2×
Ying Zhang2×
Tobias Holtdirk1×

Research Timeline

2026
Supply-Chain Poisoning Attacks Against LLM Coding Agent Skill Ecosystems

The paper introduces Document-Driven Implicit Payload Execution (DDIPE) to demonstrate that malicious code can be embedded in LLM agent skill documentation, allowing supply-chain attacks to hijack agent actions without explicit prompts.

Credential Leakage in LLM Agent Skills: A Large-Scale Empirical Study

This study conducts a large-scale empirical analysis of third-party LLM agent skills, identifying that credential leakage is a pervasive, cross-modal issue primarily caused by debug logging and resulting in exploitable, persistent secrets.

CAST: Non-Privileged Clipped Asymmetric Self-Teaching with Advantage Flipping for GRPO

The paper proposes CAST, an answer-free self-distillation method that enhances Group Relative Policy Optimization (GRPO) for verifiable rewards, allowing token-level advantage signals even when all sampled trajectories are uniformly correct or incorrect.

Automated reproducibility assessments in the social and behavioral sciences using large language models

This paper shows that large language models can automate reproducibility assessments in the social and behavioral sciences.

Highlighted terms show continued research focus across papers

Papers

cs.AIEmpiricalRecentJun 11, 2026

Automated reproducibility assessments in the social and behavioral sciences using large language models

Tobias Holtdirk, Pietro Marcolongo, Anna Steinberg Schulten, Felix Henninger +6 more

This paper shows that large language models can automate reproducibility assessments in the social and behavioral sciences.

View →
cs.AIRecentMay 29, 2026

CAST: Non-Privileged Clipped Asymmetric Self-Teaching with Advantage Flipping for GRPO

Yang Li, Gongle Xue, Yijia Guo, Yuheng Yuan +2 more

The paper proposes CAST, an answer-free self-distillation method that enhances Group Relative Policy Optimization (GRPO) for verifiable rewards, allowing token-level advantage signals even when all sa…

View →
cs.CRcs.AIcs.CLRecentApr 3, 2026

Supply-Chain Poisoning Attacks Against LLM Coding Agent Skill Ecosystems

Yubin Qu, Yi Liu, Tongcheng Geng, Gelei Deng +4 more

The paper introduces Document-Driven Implicit Payload Execution (DDIPE) to demonstrate that malicious code can be embedded in LLM agent skill documentation, allowing supply-chain attacks to hijack age…

View →
cs.CRcs.AIRecentApr 3, 2026

Credential Leakage in LLM Agent Skills: A Large-Scale Empirical Study

Zhihao Chen, Ying Zhang, Yi Liu, Gelei Deng +6 more

This study conducts a large-scale empirical analysis of third-party LLM agent skills, identifying that credential leakage is a pervasive, cross-modal issue primarily caused by debug logging and result…

View →