Browse All Jobs

Research Engineer / Research Scientist, Multimodal at Anthropic

Research Engineer/Scientist for multimodal AI systems at Anthropic in Zürich.
Hybrid
On-Site
Switzerland
Undisclosed salary

Anthropic is seeking a Research Engineer/Research Scientist to contribute to the development of reliable, interpretable, and steerable AI systems, with a focus on multimodal capabilities. The role involves research, development, and infrastructure work for state-of-the-art large language models.

The candidate will work within a multidisciplinary team of researchers, engineers, and other experts, to ensure AI is safe and beneficial.

Responsibilities include:

  • Designing training losses for new modalities
  • Designing and running experiments to evaluate scalability of architectural variants
  • Analyzing and debugging large-scale training runs
  • Scaling architectures to optimize throughput on thousands of GPUs
  • Building new deep learning architectural components from scratch
  • Building pipelines to ingest novel data sources
  • Building language model evaluations
  • Building tools for data visualization
  • Reviewing scientific literature and writing design documents

The ideal candidate should possess:

  • A Bachelor's degree in a related field or equivalent experience
  • Substantial software engineering experience
  • Research experience with scientific publications or industrial projects
  • Strong interest in the societal impacts of AI
  • Experience with high-performance, large-scale machine learning systems
  • Familiarity with ML hardware, frameworks (TPUs, GPUs, Jax, PyTorch), and infrastructure (OS internals, Kubernetes)
  • Experience with language modeling using transformers
  • Experience with deep learning research on images, videos, audio, or other modalities

Anthropic Offers:

  • Competitive compensation and benefits
  • Optional equity donation matching
  • Generous vacation and parental leave
  • Flexible working hours
  • A collaborative office space
Apply

Anthropic

Anthropic is a public benefit corporation focused on AI safety and benefiting society. Headquartered in San Francisco, it brings together researchers, engineers, policy experts, and business leaders to create reliable, interpretable, and steerable AI systems. The company's mission centers around ensuring AI is both safe and beneficial for users. Anthropic emphasizes collaborative, large-scale research with an empirical approach, drawing parallels to scientific disciplines like physics and biology. Communication and a diverse range of perspectives are valued to address AI's ethical implications.