Research Intern (bs/ms/phd)

Pika Labs Pika Labs · AI Frontier · Palo Alto, CA · Research

Pika Labs is seeking Research Interns to contribute to pre-training and mid-training of large multimodal foundation models (FMs). The role involves developing and improving methods for multimodal MLLM inferencing, with opportunities to deploy work in real-world systems. Interns will work on large-scale training efforts, design and optimize algorithms for data pipelines and inference, and collaborate on core AI technologies.

What you'd actually do

  1. Contribute directly to large-scale pre-training and mid-training efforts for multimodal foundation models.
  2. Assist in designing, prototyping, and optimizing algorithms for high-throughput data pipelines and multimodal MLLM inference.
  3. Work with diverse data types, including text, image, audio, and video, to support advanced generative model training.
  4. Collaborate closely with experienced researchers and engineers on state-of-the-art core technologies.
  5. Learn and apply frontier research techniques and production-scale AI workflows.

Skills

Required

  • Python
  • PyTorch
  • TensorFlow

Nice to have

  • machine learning
  • deep learning
  • multimodal AI
  • MLLM inference
  • generative models

What the JD emphasized

  • large-scale pre-training
  • multimodal foundation models
  • multimodal MLLM inference
  • frontier research

Other signals

  • multimodal foundation models
  • large-scale pre-training
  • MLLM inference