Research Scientist, AI Safety & Behavior Evaluation

Slope; Paris; Hybride; EUR 60 000 - 80 000; Plein temps

Avantages offerts par ce poste

Comprehensive medical insurance

Paid time off

Relocation package available

Team off-sites twice a year

Résumé du poste

White Circle is looking for a research scientist based in Paris to explore how LLM agents fail. The role encompasses owning research projects, developing audit agents to discover model behavior, and pressure-testing AI in high-stakes scenarios. A strong background in empirical research and machine learning engineering is required.

The position offers a hybrid working model, medical insurance, comprehensive support for AI tools, and team off-site retreats.

Qualifications

• Demonstrated ability to work in AI safety or related fields.
• Experience with fine-tuning and evaluating AI models.
• Ability to design experiments based on vague behavioral questions.

Responsabilités

• Own research projects end to end from idea to execution.
• Develop automated audit agents to capture AI model behavior.
• Publish findings in public forums and conferences.

Connaissances

Empirical research in agent behaviour

Strong ML engineering

Experimental design skills

Fluent with frontier AI models

Formation

MSc or PhD in a relevant field

Outils

Modern interpretability tools

Description du poste

White Circle is looking for a research scientist based in Paris to explore how LLM agents fail. The role encompasses owning research projects, developing audit agents to discover model behavior, and pressure-testing AI in high-stakes scenarios. A strong background in empirical research and machine learning engineering is required.

The position offers a hybrid working model, medical insurance, comprehensive support for AI tools, and team off-site retreats.