AI safety researcher

Lei Ma

University of Alberta

Google Scholar dblp ORCID Homepage

Last 12 months5 papers

Evaluation and benchmarks 2Attacks 1

AI safety papers from the last year

On this site: 51 papers

July 2026

Code-MUE: Measuring Code LLMs' Uncertainty through Execution-based Semantic Interaction Graphs
Jul 14arXiv ↗

April 2026

Supply-Chain Poisoning Attacks Against LLM Coding Agent Skill Ecosystems
Apr 3AttacksarXiv ↗
How Your Credentials Are Leaked by LLM Agent Skills: An Empirical Study
Apr 3Evaluation and benchmarksarXiv ↗

January 2026

From Conflicts to Collisions: A Two-Stage Collision Scenario-Testing Approach for Autonomous Driving Systems
Jan 14arXiv ↗
Evaluating Implicit Regulatory Compliance in LLM Tool Invocation via Logic-Guided Synthesis
Jan 13Evaluation and benchmarksarXiv ↗

Authorship follows dblp and the author's own Google Scholar profile; only AI safety papers on this site are listed.