Skip to main content

Senior NLP/LLM Researcher

EthosiaTel Aviv District, IsraelNot specifiedFull-timeSeniority: Senior

Posted Jun 17 · 0 applicants

Salary not listed for this role

Saving, applying or scoring takes a few seconds to set up your free account.

Willbi insight

The role in plain words

This role focuses on developing real-time protections to detect and mitigate adversarial attacks on AI agents, ensuring they do not leak personal data or hallucinate. The researcher will design advanced metrics to evaluate agent performance and conduct R&D on model protections and NLP evaluation techniques. Day-to-day tasks involve translating research findings into practical applications and collaborating with team members on model safety.

Must-have
  • MSc. / PhD in Computer Science, Mathematics, or a related data-focused field with 4 years of data science experience, OR BSc. with 7 years of data science experience
  • At least 3 years of experience in NLP
  • Proficiency in Python
  • Proficiency in PyTorch
  • Proficiency in transformers framework
Nice-to-have
  • Experience in Cyber Security Research & Red Teaming
  • Experience in LLM safety & alignment (jailbreak defenses, PII detection, content moderation)
  • Experience building, evaluating or red-teaming agentic systems

Extracted from the job description · kept up to date automatically

Who this suits

This role suits experienced researchers with a strong mathematical background, proficiency in Python and PyTorch, and at least 3 to 7 years of data science and NLP experience. It is less ideal for those without a solid foundation in computer science or mathematics, or those lacking experience in production-level product development.

Full job description

Original listing · kept for reference

Key Responsibilities:

• Create real-time protections: Develop innovative methods to detect and mitigate adversarial attacks on agents and keep them in line with the organization's policy such as avoiding hallucinations, leaking personal data etc.

• Develop new and unique evaluation methods: Implement and develop advanced metrics to evaluate agents' performance on multiple fronts.

• Research and Development: Conduct research on adversarial attacks, model protections and performance evaluation techniques for NLP models. Explore and develop state-of-the-art methods and propose innovative solutions.

• Collaboration: Work closely with other team members, participate in brainstorming sessions, and contribute to the team's success.

Qualifications:

• Professional Experience:

• MSc. / PhD in Computer Science, Mathematics, or a related data-focused field And 4 years of data science experience.

• Or BSc. in Computer Science, Mathematics, or a related data-focused field And 7 years of data science experience.

• At least 3 years of experience in NLP.

• Professional Experience Advantages:

• Cyber Security Research & Red Teaming: Experience in identifying, exploiting, and/or mitigating security vulnerabilities through hands-on research & adversarial testing.

• LLM safety & alignment: Experience developing & training LLM guardrails, such as jailbreak defenses, PII detection, content moderation etc.

• Agents & tool-using LLMs: Experience building, evaluating or red-teaming agentic systems.

• Technical Skills: Proficiency in Python and deep learning frameworks like PyTorch and transformers.

• Strong mathematical background to understand and apply complex concepts. Ability to translate research findings into practical applications.

• Excellent interpersonal skills for effective collaboration and teamwork.

• Research Experience in conducting theoretical or experimental research, using statistical tools and mathematical formulations to solve complex problems.

• Professional Experience in developing products in production environments, creating features from conception to completion.

About Ethosia
Company profile · coming soon

Employee reviews · coming soonMore roles at Ethosia

Questions about this role

  • This listing did not state a salary. We only show pay when the employer publishes it.
Ethosia
Posted Jun 17 · 0 applicants
See how you match