Skip to content
Neuer Job?Nejo!

Die KI-Suchmaschine für Jobs

ANAnthropic

Research Engineer / Research Scientist, Pre-training(m/w/x)

Zürich
ab CHF 280.000 - 680.000 / Jahr
Vollzeitmit HomeofficeBerufserfahren
AI/ML
Data Science

Developing multimodal architectures and data processing optimizers for large-scale model training. Expertise in Python, deep learning, and ML accelerators required. Optional equity donation matching and generous parental leave.

Anforderungen

  • Degree (BA required, MS or PhD preferred) in Computer Science, Machine Learning, or related field
  • Strong software engineering skills; proven track record building complex systems
  • Expertise in Python and deep learning frameworks
  • Experience with high-performance, large-scale ML systems, language modeling
  • Familiarity with ML Accelerators, Kubernetes, large-scale data processing
  • Strong problem-solving skills and results-oriented mindset
  • Excellent communication skills and collaborative work ability
  • Significant software engineering experience
  • Ability to balance research goals with practical engineering constraints
  • Willingness to take on tasks outside job description
  • Enjoyment of pair programming and collaborative work
  • Eagerness to learn about machine learning research
  • Enthusiasm for working in a cohesive team on large-scale AI research
  • Ambitious goals for AI safety and long-term progress
  • At least a Bachelor's degree in a related field or equivalent experience

Aufgaben

  • Conduct research on model architecture, algorithms, data processing, and optimizers
  • Implement solutions for model architecture, algorithms, data processing, and optimizers
  • Lead small research projects independently
  • Collaborate with team members on larger initiatives
  • Design, run, and analyze scientific experiments
  • Optimize and scale training infrastructure for efficiency and reliability
  • Develop and improve dev tooling for team productivity
  • Contribute to the entire stack, from optimization to model design
  • Optimize throughput of novel attention mechanisms
  • Propose Transformer variants
  • Experimentally compare Transformer variant performance
  • Prepare large-scale datasets for model consumption
  • Scale distributed training jobs to thousands of accelerators
  • Design fault tolerance strategies for training infrastructure
  • Create interactive visualizations of model internals

Berufserfahrung

  • ca. 1 - 4 Jahre

Ausbildung

  • Bachelor-Abschluss

Sprachen

  • Englischverhandlungssicher

Tools & Technologien

  • Python
  • deep learning frameworks
  • ML systems
  • language modeling
  • ML Accelerators
  • Kubernetes

Benefits

Attraktive Vergütung

  • Competitive compensation

Gemeinnützige Ausrichtung

  • Optional equity donation matching

Mehr Urlaubstage

  • Generous vacation

Großzügige Elternzeit

  • Generous parental leave

Flexibles Arbeiten

  • Flexible working hours

Modernes Büro

  • Collaborative office space
Die Originalanzeige dieses Stellenangebotes in der aktuellsten Version findest du hier. Nejo hat diesen Job automatisch von der Website des Unternehmens Anthropic erfasst und die Informationen auf Nejo mit Hilfe von KI für dich aufbereitet. Trotz sorgfältiger Analyse können einzelne Informationen unvollständig oder ungenau sein. Bitte prüfe immer alle Angaben in der Originalanzeige! Inhalte und Urheberrechte der Originalanzeige liegen beim ausschreibenden Unternehmen.

  • Anthropic

    Research Engineer, Production Model Post Training(m/w/x)

    Vollzeitmit HomeofficeBerufserfahren
    Zürich
  • Mistral

    AI Scientist(m/w/x)

    Vollzeitmit HomeofficeKeine Angabe
    Zürich
  • ANYbotics

    Senior AI Research Engineer, Visual Perception(m/w/x)

    Vollzeitmit HomeofficeSenior
    Zürich
  • Sony AI

    Research Intern for Deep Generative Modeling(m/w/x)

    VollzeitPraktikummit Homeoffice
    Schlieren
  • ANYbotics

    Senior AI Research Engineer in Visual Perception(m/w/x)

    Vollzeitmit HomeofficeSenior
    Zürich
Alle 100+ ähnlichen Jobs ansehen

Nejo ist eine KI – Ergebnisse können unvollständig sein oder Fehler enthalten