That's why we're building AI models and RL Environments with time awareness that reflect and evolve to make smarter decisions.
Our research intersects two topics that
will drive the next wave of AI advances
Applying causal reasoning to reinforcement learning enables AI systems that are more robust, interpretable, aware of cause-and-effect relationships and capable of making better decisions over time.
Building AI that can reflect on its own decisions and learn from experience, creating truly adaptive intelligence aligned with human values.
Generalizable Human Prior + Unbounded RL Compute + Environment
= Superintelligence.Shuchao Bi, Meta SuperIntelligence Labs
ABIDE AI RESEARCH LAB
BUILDING REFLECTIVE INTELLIGENCE
EST. 2025
CAUSAL METHODS × REINFORCEMENT LEARNING
TEMPORAL AWARENESS × HUMAN ALIGNMENT