About The Position
As part of Apple’s Multimodal Intelligence Team (MINT), the AI Research Intern will advance post-training capabilities of multimodal foundation models for agentic applications. The work includes researching and developing methods to improve how models understand, reason about, and interact with complex environments, using techniques such as supervised fine-tuning and reinforcement learning from varied reward signals. The intern will design evaluation frameworks for agent performance on realistic tasks and experiment with training strategies to improve models’ abilities to perceive multimodal inputs, understand intent, and execute complex autonomous behaviors. The role involves working with large-scale datasets combining visual, textual, and interaction data, and keeping current with emerging research in foundation-model agents to identify techniques that could advance autonomous systems. The intern will collaborate with researchers and engineers on cutting-edge research, with the primary goal of producing novel work suitable for publication at a top-tier conference. The position is available for 6–12 months.
Requirements
- Currently pursuing a PhD degree or equivalent experience in Machine Learning, Computer Vision, Natural Language Processing, Data Science, Statistics, or a related area.
- Experience with large language models or vision-language models and their application in agentic systems.
- Proficient programming skills in Python.
- Experience with at least one modern deep learning framework: PyTorch, JAX, or TensorFlow.
- Demonstrated publication record in relevant conferences such as NeurIPS, ICML, ICLR, or CVPR.
- Experience with foundation models, including language, vision-language, or multimodal models.
- Experience with post-training (SFT or RL) to optimize large models for agentic systems.
- Available for a 6–12 month internship.
Responsibilities
- Research and develop post-training methods for multimodal foundation models used in agentic applications, including supervised fine-tuning and reinforcement learning from various reward signals.
- Design evaluation frameworks to assess agent performance on realistic tasks.
- Experiment with training strategies to improve models’ multimodal perception, intent understanding, reasoning, and autonomous behavior.
- Work with large-scale multimodal datasets combining visual, textual, and interaction data.
- Stay current with research in foundation-model agents and identify techniques that could advance autonomous systems.
- Collaborate with researchers and engineers on cutting-edge research.
- Aim to produce novel research suitable for publication at a top-tier conference.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
What This Job Offers
Career Level
Intern
Education Level
Ph.D. or professional degree
