ML Research & Engineering Internship

Technical Staff
Paris / New York

About the team

Adaptive ML is helping companies build singular generative AI experiences by democratizing the use of reinforcement learning. We are building the foundational technologies, tools, and products required for models to learn directly from users' interactions and for models to self-critique and self-improve from simple written guidelines. Our tightly-knit team was previously involved in the creation of state-of-the-art open-access large language models such as Falcon-180B. We have closed a $20M seed with Index & ICONIQ, and are looking forward to shipping a first version of our platform, Adaptive Engine, in early 2024.

Our Technical Staff is responsible for building the foundational technology powering Adaptive ML, in line with requests and requirements identified by our Product and Commercial Staff. We strive to build excellent, robust, and efficient technology, and to conduct at-scale, honest research with high-impact for our roadmap and customers.

About the role

This is an internship for students enrolled in an M.Sc. or Ph.D. We strongly favour longer internships of 5-6 months, and typically require for interns to be in-person at either our Paris or New York office. Our ideal outcome for an internship is to convert into a full-time role.

As a an ML Research & Engineering Intern in our Technical Staff, you will be in contact with the foundational technology powering Adaptive ML, typically by working with our internal LLM Stack, Adaptive Harmony. We fundamentally believe that generative AI is best approached as a so-called big science combining large-scale engineering and stringent empirical research. Accordingly, we have a strong bias for doing things at-scale, and for systematic empirical demonstrations.

Some examples of tasks interns in ourTechnical Staff may pursue on a daily basis, with the assistance of their supervisor:

  • Develop robust software in Rust, interfacing between easy-to-use Python recipes and high-performance distributed training code running on hundreds of GPUs;
  • Profile and iterate GPU inference kernels in Triton or CUDA, identifying memory bottlenecks and optimizing latency—and decide how to adequately benchmark an inference service;
  • Develop and execute an experiment plan for better understanding the nuances between DPO and PPO in a fair and systematic way;
  • Build data pipelines to support reinforcement learning from noisy and diverse users' interactions across varied tasks;
  • Experiment with novel ways to combine adapters to steer the behaviour of language models;
  • Build hardware correctness tests to identify and isolate faulty GPUs at scale.

We are looking for self-driven, intense individuals, who value technical excellency, honesty, and growth.

Your responsibilities

Generally,

  • Lean from and contribute to the foundational technology powering Adaptive ML, with a focus on high-performance software engineering and large-scale RL research;
  • Identify promising trends and high-impact findings from the litterature;
  • Report clearly on your work to a distributed collaborative team, with a bias for asynchronous written communication.

On the engineering side,

  • Write high-quality software in Rust, with a focus on performance and robustness;
  • Profile dedicated GPU kernels in CUDA or Triton, optimizing across latency/compute-bound regimes for complex workloads;
  • Identify and resolve bugs in large distributed systems, at the intersection of software and hardware correctness.

On the research side,

  • Conduct research on large language models or diffusion models, systematically exploring how reinforcement learning can be used to personalize models;
  • Reproduce results from the RL, LLM, and diffusion literature, separating the fluff and the groundbreaking;
  • Learn to own a research agenda, with a bias for at-scale, systematic empirical research.

Nearly all members of our Technical Staff hold a position that is a blend of engineering and research.

Your (ideal) background

The background below is only suggestive of a few pointers we believe could be relevant; we welcome applications from candidates with diverse backgrounds, do not hesitate to get in touch if you think you could be a great fit even if the below doesn't fully describe you.

  • Pursuing a M.Sc./Ph.D. in computer science, or demonstrated experience in software engineering, preferably with a focus on machine learning;
  • Strong programming skills, especially regarding distributed problems where performance is key;
  • Contributions to relevant open-source projects, such as efficient implementations of models and RL;
  • Previous publications at top-tier machine learning venues (e.g., NeurIPS, JMLR);
  • Passionate about the future of generative AI, and eager to build foundational technology to help machines deliver more singular experiences.

Benefits

  • Competitive salary and assistance relocating to Paris/New York for the duration of the internship;
  • Mentorship from industry experts on large language models, and a network of academy and industry connections for collaborations;
  • Paid travel to conferences for publications during the internship;
  • A full-time offer at the end of the internship if there is a fit.

Apply for this position.

Send us an e-mail with your resume.

Learn more about Adaptive ML.