Justin Svegliato
Scholar

Justin Svegliato

Google Scholar ID: 3Orv6wUAAAAJ
Senior Research Scientist, Microsoft / UC Berkeley
Artificial IntelligencePlanningReinforcement LearningLarge Language ModelsAI Safety
Citations & Impact
All-time
Citations
775
 
H-index
13
 
i10-index
21
 
Publications
20
 
Co-authors
20
list available
Resume (English only)
Academic Achievements
  • Received a Distinguished Paper Award at AAAI.
  • Awarded the NSF Graduate Research Fellowship.
  • Finalist for the Distinguished Teaching Award at UMass Amherst.
  • Published numerous papers at top-tier conferences including NeurIPS, ICLR, ICRA, IROS, AAAI, AAMAS, IJCAI, and ICAPS.
  • Notable works include AssistanceZero (a method for solving assistance games), Tensor Trust (an LLM-based game and prompt injection dataset), metareasoning for safe decision-making, ethically compliant autonomous systems, LLM calibration, and active reward learning.
  • Published an article in Artificial Intelligence Journal (AIJ) on competence-aware systems.
  • Co-organized workshops on AI safety, including the CHAI Workshop and the IROS Workshop on Ethical Robotic Systems.