Research Intern, Efficient Deep Learning

This is a fully remote position, open to applicants in California.

📋 Description

• Investigate, design, and create innovative techniques for efficient deep learning in diffusion LLMs and multimodal models.

• Explore efficient agentic AI utilizing hybrid inference orchestration across both cloud and edge environments.

• Focus on enhancing sampling efficiency, adaptive unmasking, self-speculation/parallel decoding, training and distillation pipelines, as well as multimodal generation.

• Develop strategies for routing and scheduling policies, including on-device versus cloud expert delegation, and resource-aware agent loops.

• Contribute to the publication of original research findings.

• Collaborate effectively with team members and other teams.

• Engage with product groups to facilitate technology transfer.

• Work alongside external researchers for collaborative projects.

⛳️ Requirements

• Currently pursuing a Ph.D. in Computer Science/Engineering, Electrical Engineering, or a related field.

• Strong understanding of the principles and practices of machine learning and deep learning.

• Required experience with large language models, diffusion language models, multimodal/vision-language models, or agentic systems.

• Practical experience in large-scale model training, including data preparation and model parallelization (tensor and pipeline), is essential.

• Proven research track record with at least one publication in a top-tier conference (ICML, ICLR, NeurIPS, CVPR, ICCV, etc.).

• Excellent communication abilities.

• Experience in parallel programming (e.g., CUDA).

• Interest or experience in hybrid cloud-edge inference, orchestration, or adaptive routing.

• Background in pruning, quantization, NAS, or efficient model architectures.

🏝️ Benefits

• Competitive salaries.

• Comprehensive benefits package.

• Intern benefits.