Skills

Communication Python GitHub Research Training Architecture Machine Learning PyTorch Deep Learning Programming C++

Job Specifications

IMPORTANT NOTE: Numenta is unable to consider internship applications from international applicants unless they are currently enrolled in a US University, are US Citizens, or hold a Green Card or are NAFTA (Mexican or Canadian) citizens.

Numenta is looking for outstanding machine learning researchers for the summer of 2026 internship opportunities. If you know how LLMs work and love diving into the details in search of better techniques, you’ve come to the right place!

The commitment is for 3 to 4 months full-time (longer appointments up to one year may be possible). Numenta’s interns will participate in cutting-edge machine intelligence research and development. The ideal candidate will be an excellent programmer with hands-on experience in training and evaluating machine learning systems.

The position is based in Menlo Park, California.

What You Can Expect to Learn and How You Might Contribute

Understand the challenges of implementing performant deep learning models.
Accelerate neural network architectures through biologically inspired sparse algorithms.
Improve your software engineering skills in a collaborative environment.
Implement and optimize core learning algorithms.
Evaluate and tune current systems for both speed and accuracy.

Desired Qualifications

Currently enrolled in an MS or PhD program, or a recent graduate, in Computer Science or a related field.
Outstanding programming skills in Python, especially PyTorch. Expect to work with FSDP / distributed training and be comfortable with PyTorch hooks and writing custom modules.
Deep understanding of machine learning systems, especially LLMs. For example, you are familiar with tokenizers, the transformer architecture, common training protocols, and common inference algorithms.
Capable of analyzing machine learning systems and identifying weaknesses.
Experience with sparsity and sparse learning algorithms is a plus.
Experience with writing CUDA code, Triton kernels, or writing performant C++ code are pluses.
Able to demonstrate strong intellectual capabilities and significant individual contributions.
Excellent written and verbal communication skills.

Compensation:

This is a paid, temporary intern position. The salary ranges from $2,600 to $3,400 paid weekly. The exact salary will depend on the applicant’s academic experience. Since this is a temporary position, other benefits (medical, dental, stock options, housing or travel reimbursements, etc.) are not offered.

How To Apply:

Qualified applicants should send the following:

Cover letter describing their specific interest in Numenta.
Resume, including a list of relevant coursework and a link to your GitHub repository or other public code repository that contains code you have worked on.

Please email your resume and cover letter to interns@numenta.com.

Numenta is an equal opportunity employer supporting workforce diversity.