
Cohere For AI Scholars Program

Scholars Program
The goal of the Scholars Program is to create state-of-the-art research on frontier AI, and to provide a door to a top-tier industry research lab for researchers at the beginning of their careers.
Our research scholars become part of a dedicated team of passionate researchers and industry experts. This experience, pairs scholars with a project proposal and top-tier researchers, to facilitate growth as a researcher. The program entails full-time, paid participation, providing scholars access to a large-scale experimental framework and world-class research mentors.
About the program
Since 2023, we’re proud to have welcomed three cohorts to the Cohere For AI Scholars Program to help close the gap between research experience and opportunity. We have supported talented researchers from around the world to work on open-ended questions with our in-house research team. We are incredibly proud of the many scientific contributions these scholars have made during the program.
Machine learning is advancing rapidly, but limited resources restrict fundamental NLP research and large-scale ML experiments. To foster new developments, it is crucial to broaden access and encourage participation in fundamental research. Our mission is to address these gaps and empower the next generation of rising stars as they venture into their research endeavors.

Since 2023, we’ve been proud to run three cohorts of the Cohere For AI Scholars Program to help close the gap between research experience and opportunity. In this time, we have welcomed talented researchers from around the world to explore open-ended questions with our in-house experienced research team.
Since 2023, we’ve been proud to run three cohorts of the Cohere For AI Scholars Program to help close the gap between research experience and opportunity. In this time, we have welcomed talented researchers from around the world to explore open-ended questions with our in-house experienced research team.
3
Cohorts of Scholars to Date
18
Countries Represented Among Scholars
14
Published Papers
25+
Project Mentors
Papers Published by Our Scholars
RLHF Can Speak Many Languages: Unlocking Multilingual Preference Optimization for LLMs
We introduce a novel, scalable method for generating high-quality multilingual feedback data to balance data coverage. We establish the benefits of cross-lingual transfer and increased dataset size in preference training.
LLM See, LLM Do: Guiding Data Generation to Target Non-Differentiable Objectives
Our work exhaustively characterizes the impact of passive inheritance of model properties by systematically studying the consequences of synthetic data integration.
The Multilingual Alignment Prism: Aligning Global and Local Preferences to Reduce Harm
We explore the viability of different alignment approaches when balancing dual objectives: addressing and optimizing for a non-homogeneous set of languages and cultural preferences while minimizing both global and local harms.
Aya 23: Open Weight Releases to Further Multilingual Progress
Our technical report shares evaluation results on multiple multilingual NLP benchmarks, and generation quality assessments.
To Code, or Not To Code? Exploring Impact of Code in Pre-training
Including code in the pre-training data mixture, even for models not specifically designed for code, has become a common practice in LLMs pre-training. e ask "what is the impact of code data used in pre-training on a large variety of downstream tasks beyond code generation".
Multilingual Arbitrage: Optimizing Data Pools to Accelerate Multilingual Progress
Can you surpass individual model performance by sampling parts of the distribution strategically from a pool of models?
We introduce “multilingual arbitrage” to describe capitalizing on performance variations to produce large gains in performance.
Nexus: Specialization meets Adaptability for Efficiently Training Mixture of Experts
In our latest work, we ask “How to make Mixture-of-Expert models more efficient, specialized, and adaptable at once?” We introduce Nexus, which enables specialization and adaptability within the efficient upcycling framework.
From One to Many: Expanding the Scope of Toxicity Mitigation in Language Models
Work to date on toxicity mitigation has exclusively focused on the English language. Yet harm exists in all languages. We delve into the complexities of toxicity assessment across 9 languages spanning 5 scripts, ranging from high to mid-resource availability.
Learn More
Scholars Program: Research Journeys Start Here
Blog post sharing details about our most recent recruitment cycle.
Reflecting on the journey
2024 Scholar, Luísa Shimabucoro, reflects on her experience
Article in Fortune announcing the Scholars Program
Expanding access might be the key to resolving the A.I. talent shortage
Frequently Asked Questions
When can I apply for the Scholars Program?
Applications open each year in August; the Scholars Program begins annually in January.
Do I need previous research experience to apply to the Scholars Program?
No. In fact, we are particularly interested in identifying candidates with strong engineering skills and demonstrated creative thinking but limited experience or no experience with published papers. You need not have to have previous research experience to apply.
Do I need to submit a research project proposal work on during the Scholars Program?
We want you to focus on showcasing your skills in the application process, so there is no need to propose a research project. We will match our Scholars to research projects and mentor
How does project matching work?
We will consider the skills, interests, and background of our successful applicants alongside our research projects to make the best possible match between Scholar and project.
Is the Scholars Program paid?
Yes, this is a full-time, paid position.
Is the Scholars Program remote, or will I need to relocate to a Cohere office?
The Scholars Program will operate on a remote-first basis. We want to make this opportunity available to emerging ML researchers around the world. If you are based out of or visiting a city with a Cohere office (Toronto, San Francisco, New York or London), you are welcome to come work with us in person.
Will C4AI sponsor work permits?
The Scholars Program is structured as a remote-first opportunity, with no requirement to relocate.
Is it possible to do the Scholars Program on a part-time basis?
01:
The Scholars Program is structured as a rigorous full-time commitment. In an effort to keep up with the pace of advancement of ML research, we look forward to supporting Scholars dive into the research process as they work towards publication within the program period.
I was unsuccessful in my application–can I reapply for future Scholars Programs?
Yes–we welcome re-applications for future cohorts. We see this as a positive signal of your continued interest and commitment to the program.

Cohere For AI
Cohere For AI is Cohere's non-profit research lab that seeks to solve complex machine learning problems. We support fundamental research that explores the unknown, and are focused on creating more points of entry into machine learning research.