Lab V2

Seeding the Next Generation of Artificial Intelligence and Machine Learning

Follow on Twitter @lab_v2

GitHub site: lab-v2

Check out our new YouTube channel dedicated to
Neuro Symbolic AI.

Interested in joining Lab V2? Read this information first.

Check out our related site dedicated to neuro symbolic AI.

Lab V2 is not hiring new Ph.D. students at this time.  We will revisit the hiring of Ph.D. students no earlier than 2025.

Society is now starting to enjoy the results of years of research into artificial intelligence.  However, major challenges remain.  Today’s state-of-the-art methods tend to function as “black boxes”, do not identify causal linkages, require vast amounts of training data, and may have unexpected poor performance in certain conditions. This leads to challenges in verifying performance, avoiding bias, explaining model outcomes, and assessing risk. Our depth of experience in hybrid symbolic-machine learning systems including programs like IARPA CAUSE, ICARUS, ONR NEPTUNE, and several awards from ARO, AFOSR, as well as real-world transitions of AI/ML systems for military, law enforcement, and cybersecurity positions Lab V2 to address these challenges.

At Lab V2, we are focused on several critical challenges in the field, including:

  • Neuro Symbolic AI (NSAI) topics including induction, fixpoint-based deduction, and causality
  • Metacognitive AI topics including performance verification and reasoning about AI and ML systems
  • Symbol grounding, i.e. the translation from vectors to symbols
  • Reasoning about agent courses of action and deception actions in geospatial settings
  • Analysis of graphical representations of first order knowledge with applications to the global supply network (GSN)
  • Applications to intelligence analysis

Lab V2 is a new research group at Arizona State University led by Paulo Shakarian.  Shakarian is a tenured professor at Arizona State who has recently returned to his position after exiting the machine learning startup he co-founded.  In Lab V2, the team will apply scientific ideas at the intersection of machine learning and symbolic AI to improve intelligent systems across a range of domain problem areas including autonomy, supply network analysis, and medical applications.