ML Research Intern

Modal Modal · Data AI · New York, NY · Engineering

Seeking PhD research interns to work on AI infrastructure, focusing on reinforcement learning, foundation models (LLMs, multimodal), large-scale model training, optimization, inference, and extending models to long-context/horizon tasks. The role involves improving existing methods and developing new techniques for efficiency, reliability, and robustness in real-world deployments.

What you'd actually do

  1. improving existing methods and developing new techniques for large-scale model training, optimization, and inference
  2. extending models to long-context and long-horizon tasks
  3. improving inference-time efficiency, reliability, and robustness in high-stakes real-world deployments

Skills

Required

  • reinforcement learning
  • machine learning
  • foundation models
  • large language models
  • multimodal models
  • large-scale model training
  • model optimization
  • model inference
  • long-context tasks
  • long-horizon tasks
  • inference-time efficiency
  • reliability
  • robustness
  • distributed training
  • multi-GPU environments
  • programming skills
  • engineering skills

Nice to have

  • experience developing and evaluating large-scale models or machine learning systems
  • familiarity with distributed training, large-scale inference, or multi-GPU environments
  • publications at leading venues such as NeurIPS, ICML, ICLR, CVPR, CoRL, UAI, JMLR, or TMLR
  • collaborative, mission-driven mindset
  • ability to work effectively across research and engineering teams

What the JD emphasized

  • strong research experience in reinforcement learning, machine learning, and foundation models
  • demonstrated record of research in reinforcement learning, machine learning, foundation models, or related areas
  • Publications at leading venues such as NeurIPS, ICML, ICLR, CVPR, CoRL, UAI, JMLR, or TMLR

Other signals

  • building AI infrastructure
  • improving existing methods and developing new techniques for large-scale model training, optimization, and inference
  • extending models to long-context and long-horizon tasks
  • improving inference-time efficiency, reliability, and robustness