PhD Research Scientist Intern - Multimodal LLM & S2ST

Meta; Redmond; On-site; AUD 56,000 - 84,000; Full time

Job summary

Meta Reality Labs Research invites interns to advance multimodal large language models and speech-to-speech translation for AR/VR experiences. Interns will explore models integrating audio, speech, vision, and language, and collaborate with researchers to showcase real-world potential.

Internship durations range from 12 to 24 weeks with multiple start dates; applicants should be pursuing a PhD in a related field and have strong Python and ML framework experience, plus a track record of

Qualifications

• PhD candidate or in progress in a STEM field relevant to ML/AI.
• 3+ years Python programming experience.
• 3+ years with PyTorch or TensorFlow.
• 2+ years in speech, NLP, multimodal or audio-visual learning.
• Must have work authorization at hire and during employment.
• Experience with multimodal LLMs and speech translation is preferred.
• Publications in top venues (NeurIPS/ICML/ACL/CVPR/etc.).
• Strong software engineering, internships, or OSS contributions.
• Ability to communicate across disciplines.
• Intent to return to degree program after internship.

Responsibilities

• Investigate research questions in multimodal LLMs and S2ST.
• Develop and evaluate novel models across audio, speech, vision, and language.
• Collaborate with researchers and engineers to demonstrate applicability.

Skills

Python programming

ML frameworks (PyTorch/TensorFlow)

Speech/NLP/Multimodal models

Work authorization

Multimodal LLM research

Speech translation

Speech foundation models

Distributed training

Model evaluation

Publications

Open-source contributions

Cross-disciplinary collaboration

Pursue degree after internship

Education

PhD in Computer Science / AI / ML / EE / Speech

Job description

Meta Reality Labs Research invites interns to advance multimodal large language models and speech-to-speech translation for AR/VR experiences. Interns will explore models integrating audio, speech, vision, and language, and collaborate with researchers to showcase real-world potential.

Internship durations range from 12 to 24 weeks with multiple start dates; applicants should be pursuing a PhD in a related field and have strong Python and ML framework experience, plus a track record of