Research Engineer/Scientist(all levels), Efficient Models

What you'll do

• Develop efficient algorithms and architectures for generative and multimodal models using methods such as distillation and quantization.
• Advance diffusion and autoregressive modeling approaches focused on acceleration and efficiency.
• Develop methods and infrastructure to transfer foundation-model capabilities into smaller, more efficient models.

What they ask for

• B.S. in Computer Science or a related field, or equivalent experience.
• Expertise in efficient models, computational bottlenecks, and acceleration methods.
• Proficiency training generative AI or LLM models with frameworks such as PyTorch and JAX.
• Strong communication and collaboration skills.

Nice to have

• Ph.D. in GenAI, MLSys, or equivalent experience.
• Research experience in GenAI, MLSys, or LLMs, with expertise in at least one listed area such as image/video generation, model compression, efficient architectures, or reinforcement learning methods.

Tools

PyTorch, JAX, diffusion, autoregressive models, MoE, window attention, RLHF, DPO

AI focus

Research and develop efficient large-scale generative AI and multimodal models, including distillation, compression, quantization, acceleration, and image/video generation.