Ai2Berkeley, CAPostdoctoral researcher leading high-impact AI projects on Mixture-of-Experts and long-context language models, training/releasing models, building open-source tools, publishing papers, and mentoring juniors. Requires recent PhD in CS/ML with strong publication record and PyTorch expertise.
As a Research Scientist II on the Video team, you will drive core research initiatives, deliver reproducible experimental results, and help translate machine learning models into real-world product solutions, focusing on real-time video processing and deepfake detection.
Develops and optimizes GPU-accelerated kernels and algorithms for ML/AI applications, co-designing with modeling, hardware, and software teams. Requires strong GPU programming expertise in CUDA/Triton and knowledge of ML models.
Develops scalable ML-based planning and prediction systems for autonomous driving trajectories. Requires expertise in sequential decision making, deep RL, imitation learning, generative modeling, and robotics; M.Sc./Ph.D. preferred with top conference publications.
Deploys high-performance vision and multimodal AI models to robotic platforms, optimizing for low-latency inference under hardware constraints. Requires 3+ years in real-time robotics systems, expertise in PyTorch, C++/Rust, and edge deployment.
Conducts research on foundation models for robotics, focusing on manipulation, sim-to-real transfer, RL, and skill learning using simulations and real robots. Requires PhD-level expertise in robotics/AI, strong publications, and software skills for open-source breakthroughs.