Research Scientist Intern, Multimodal AI (PhD)

Meta is a leading technology company focused on connecting people through innovative solutions such as augmented and virtual reality. They are seeking a Research Scientist Intern to contribute to groundbreaking research in audio signal processing and machine learning for AR/VR applications, working alongside a talented team to develop novel algorithms and evaluation protocols.

Responsibilities:

• Design, implement, and maintain comprehensive evaluation protocols for large language models, including both automated and human-in-the-loop assessments
• Develop and curate high-quality datasets and benchmarks to measure model performance, safety, fairness, and robustness across a variety of tasks and modalities
• Analyze model outputs to identify strengths, weaknesses, and failure modes, and provide actionable insights to research and engineering teams
• Design and implementation of novel algorithms to solve audio research problems
• Collaboration with teams building Meta’s language AI products
• Collaborate with researchers, engineers, and cross-functional partners to define evaluation goals, communicate findings, and drive improvements in model quality
• Develop tools and infrastructure to streamline and scale evaluation processes, including dashboards, annotation platforms, and reporting systems
• Stay up-to-date with the latest research in audio LLM evaluation, benchmarking, and responsible AI, and incorporate best practices into Meta’s workflows
• Disseminate evaluation results through internal reports, presentations, and, when appropriate, external publications