Bell Labs Internship on Source-aware Language Models (PhD) - Remote

Overview

This internship focuses on enhancing the confidence of users in Large Language Models (LLMs) by determining the sources of information provided in their answers, thereby improving accountability and reducing misinformation.

In Short

• Work on tracking the source of knowledge in LLMs.
• Refine, implement, and test techniques for source tracking.
• Research new domains for evaluating source-aware LLMs.
• Flexible duration of 3-4 months.
• Starting time is flexible.
• Location options include Antwerp (Belgium) or Stuttgart (Germany).
• Opportunity to work in a hybrid environment.
• Gain experience in AI and LLM training methods.
• Contribute to improving user confidence in AI outputs.
• Engage in practical research and development tasks.

Requirements

• Familiarity with training of LLMs.
• Understanding of information encoding in AI.
• Ability to implement new training methods.
• Research skills to explore new application domains.
• Interest in AI and machine learning.
• Strong analytical skills.
• Good communication skills.
• Ability to work independently and in a team.
• Flexibility in working hours and tasks.
• Willingness to learn and adapt.

Benefits

• Gain hands-on experience in AI development.
• Work in a collaborative environment.
• Flexible working hours.
• Opportunity to contribute to innovative projects.
• Networking opportunities within the industry.
• Potential for future employment.
• Mentorship from experienced professionals.
• Access to resources and tools for research.
• Enhance your resume with relevant experience.
• Work in a hybrid model.