Member of Technical Staff Training

You will develop diffusion-language-model architectures, implement training objectives and loss functions, research controlled generation and multimodal methods, improve training and inference efficiency, and develop post-training techniques that improve model behavior.

Responsibilities

• Design, develop, and optimize diffusion-based language model architectures
• Implement training objectives and loss functions for discrete diffusion LLMs
• Research and implement controlled text generation and constraint-satisfaction techniques
• Develop multimodal integration methods
• Improve model efficiency, training time, and inference throughput
• Develop post-training techniques to align and improve model behavior

Requirements

• At least 2 years of experience on ML projects using PyTorch or equivalent
• Familiarity with transformers and core LLM concepts
• Familiarity with diffusion-model training and inference
• Experience training deep learning models at scale in distributed computing environments

Benefits

• Equity
• Flexible vacation and paid time off
• Health, dental, and vision insurance
• 401k match
• Catered meals
• Commuter subsidies