About this role
Conduct research on reinforcement learning algorithms for multimodal models (including diffusion and autoregressive models), design and develop RL training frameworks and reward modeling strategies for efficient large-scale training, and explore next-generation reinforcement learning paradigms for learning from environmental feedback.
Skills for this role
Reinforcement LearningMultimodal ModelsDiffusion ModelsAutoregressive ModelsImage GenerationVideo GenerationUnified Multimodal FrameworksReward ModelingReinforcement Learning Training FrameworksDeep Learning System ImplementationModel TrainingInference OptimizationCPU/GPU AccelerationDistributed TrainingText-to-Image GenerationText-to-Video GenerationResearch Publications