Chain-of-Thought Reasoning
Mixture-of-Experts
Diffusion-LLMs
Diffusion-based Trajectory Decoding
Closed-loop Training
Off-policy RL
Online RL
Enforcing Consistency
Vision-language models
Multimodal reasoning
Spatial Multimodal Models
Modality Alignment
Model Scaling
Synthetic data
Inference Optimizations (e.g., parallel decoding, speculative decoding)
Token Representations
Model Distillation
Digital Twins
Scenario Generation
World Models
Behavior/Traffic Modeling
Mapless driving
World Representations
Beyond imitation learning
Safety-aware end-to-end models
Multi-modal sensor fusion
2D/3D detection, segmentation, depth estimation, scene understanding, neural representations
Principled robustness
Model explainability
Control Barrier Functions (CBFs)
Verification and Validation of Safety-Critical AI Systems
Trustworthy AI/ML for autonomy and robotics
Data-Strategies for AI
Benchmarking AV
Click here to learn more about NVIDIA, our early talent programs, benefits offered to students and other helpful student resources related to our latest technologies and endeavors.