Research Scientist – Multimodal Foundation Models

What you'll do

• Develop vision-language models and multimodal foundation-model architectures, including post-training and reinforcement learning.
• Improve model performance, reliability, training efficiency, and data quality through experiments and optimization.
• Build evaluation frameworks and benchmarks, and conduct large-scale research using distributed training infrastructure and datasets.

What they ask for

• Hands-on experience researching, training, fine-tuning, or post-training large-scale foundation models, particularly vision-language or multimodal models.
• Strong understanding of deep learning architectures, optimization, and modern model training; proficiency in Python and PyTorch, JAX, or comparable frameworks.
• Practical experience with supervised fine-tuning, preference optimization, or reinforcement learning approaches such as DPO, GRPO, PPO, or RLHF.
• Experience with large-scale machine-learning experiments, distributed training systems, GPU-based model development, and open-weight multimodal foundation models; research publications, open-source contributions, or experience at leading AI research organizations.

Nice to have

• Experience with large-scale model post-training and reinforcement learning for reasoning or alignment.
• Familiarity with multimodal data pipelines, reward modeling, synthetic data generation, or advanced model evaluation.

Pay

$185K – $210K (base salary)

Tools

Python, PyTorch, JAX, DPO, GRPO, PPO, RLHF

AI focus

Research and development of vision-language and multimodal foundation models, including training, post-training, reinforcement learning, multimodal reasoning, evaluation, and optimization.