Salesforce AI Research Grant

Thank you for your interest in the Salesforce AI Research Grant. Our 2020 grant applications are now closed. Please check back on our website for more partnership opportunities or follow us on Twitter @SFResearch. We look forward to contacting our grant winners mid November 2020 with a blog to announce winners publicly January 2021!

Salesforce believes that it is our responsibility to further equality for all, especially as it relates to the field of AI. We apply our research to develop AI products that you can trust and solutions that benefit everyone. As we see great progress in AI and its applications, we’re looking for diverse individuals with innovative ideas who can join us in improving the state of the world.

Our Salesforce Research team is inviting submissions from university faculty, non-profit organizations, and NGOs to apply for our Salesforce AI Research Grant. Our goal is to support individuals who extend, use or analyze AI and its applications. We encourage people to submit problems or research pertaining to our current research projects but it is not required. The purpose of the grant is to create lasting relationships with our grant winners and to advance the state of the art in AI. We would like to extend this opportunity to the entire research community but this year, we plan to focus more on groups who are underrepresented in the field of AI.

Salesforce will fund up to $50,000 USD, depending on the needs to support the research. We will fund up to four different proposals.

  • AI for Good
  • AI for Economics
  • AI for Software Engineering
  • Biomedical AI
  • De-biasing Deep Learning Models
  • Democratizing Scalable Deep Learning
  • Efficient Training Algorithms
  • Ethics in AI
  • Few-Shot Learning
  • Interpretability and Robustness
  • Lifelong Learning
  • Machine Learning for AIOps
  • Multi-Task Learning
  • Natural Language Processing
  • Recommendation
  • Reinforcement Learning
  • Self-Supervised Learning



Who is eligible to apply for the research grant?

Permanent faculty members, non-profit organizations, and NGOs around the world.

General Grant Questions

Which Research topics are you most interested in receiving a grant proposal for?
(preferred but not required)
AI for Good
AI for Economics
AI for Software Engineering
Biomedical AI
De-biasing Deep Learning Models
Democratizing Scalable Deep Learning
Efficient Training Algorithms
Ethics in AI
Few-Shot Learning
Interpretability and Robustness
Lifelong Learning
Machine Learning for AIOps
Multi-Task Learning
Natural Language Processing
Reinforcement Learning
Self-Supervised Learning
Does the grant need a mentor or leader?
Yes, your submission should have one lead who will act as the point of contact.
Can I submit more than one grant proposal?
Who should write the grant?
The people applying and/or a grant writer from your organization.
Can my proposal include the 3rd page for references?
Who will own the intellectual property of the research conducted using the grant?
The Salesforce Research Deep Learning Grant is not subject to any intellectual property (IP) restrictions. IP developed solely by the grant recipient will be owned by the grant recipient or his/her organization. If the grant recipient uses the grant to conduct joint research with Salesforce, IP that is jointly developed with Salesforce will be subject to the terms of a separate research collaboration agreement.
How will grants be judged and who will review the proposals?
Grants will be judged based on the specifics noted above, the quality of the proposed research and how closely it matches the research we are currently doing at Salesforce. We will also be looking for diversity in backgrounds and areas of interest to those of the other grant winners. The proposals will be reviewed and judged by a panel of research scientists, managers and directors within the Salesforce Research team.
When will I be notified of the acceptance/decline of my proposal?
We expect to notify grant applicants by Mid November 2020.

The Proposal


What can the funding be used for?
The funding can be used to support the costs of your proposed research, including:
- Cloud computing costs (AWS, GCP, Azure, etc.)
- Indirect/overhead costs
- Student stipends
- Advertising/marketing costs
- Tuition
- Travel expenses to present your work
- Office supplies and expenses
Where is the funding coming from?, inc. Salesforce is a for-profit entity.
How many people can share a single $50,000 grant?
If my proposal is selected, when will I receive the grant funding?
Winners of the Salesforce Deep Learning Grant will be contacted by mid November 2020. We plan to have all payments made by the end of January 2021.
Chenhao Tan, University of Colorado Boulder
AI for Good, De-biasing Deep Learning Models, Explainable AI (XAI), Natural Language Processing
Active Learning with Explanations for De-biasing NLP Models

When a model makes a mistake, it ought to be possible to understand the reasoning underlying that mistake and make corrections at the level of that reasoning. Frequently-discussed issues like implicit model bias and vulnerability to adversarial attacks can be described as issues of attribution β€” models assigning spurious association to input tokens. Explainable AI offers methods for describing these lapses in reasoning, but there exists no method for correcting them efficiently. To address this gap, we propose to develop an active learning approach by selectively soliciting corrections to model reasoning at the token level.
Our approach combines insights from explainable AI and active learning to 1) efficiently sample useful examples, 2) explain model reasoning to users, 3) solicit alternative explanations via token-level labels, and 4) incorporate those corrections back into the model as additional training signal. The proposed work will enable a way to refine NLP models more directly and efficiently than traditional methods.

Christopher RΓ©, Stanford University
De-biasing Deep Learning Models, Lifelong Learning, Data Augmentation and Robustness
Muscaria: Model Patching with Weakly Supervised Data Augmentation

Data augmentation is a common technique for dramatically improving the generalization of machine learning models and has been of paramount importance to the success of computer vision. Augmentation strategies in use today have been developed as a consequence of trial and error over many years of practice, and perform generic, rather than semantic manipulations to the data. The goal of our research is to enable users to perform model patching: improving deployed models using customized, complex and semantic data augmentation pipelines spun up with weak user supervision. This work will build on our lab’s expertise in developing theory, algorithms, and systems for incorporating weak supervision into machine learning models over the last few years.

Greg Durrett, University of Texas at Austin
Natural Language Processing
Exerting Fine-Grained Content Control for Summarization

While recent neural summarization methods have achieved strong performance on standard datasets, saturating ROUGE scores in settings like CNN/Daily Mail, summarization as a whole is still far from solved. A summarization system should be usable by a range of users with different information needs across a variety of text domains, which is not the case for current systems. We propose a summarization model where users can exert fine-grained control over both content selection and generation, manipulating the system's behavior on a particular instance and, more importantly, being able to declaratively specify how to produce a summary. Our model decouples the content selection and generation processes, selecting content at a sub-sentence level with what we call an information skeleton, then generating the summary abstractively. The user can express preferences via key phrases about what information is selected in the skeleton. We propose several targeted evaluations to establish our model's overall effectiveness and controllability in our settings of interest.

Hung-yi Lee, National Taiwan University
Lifelong Learning, Multi-Task Learning, Natural Language Processing, DecaNLP
Lifelong Language Learning

The goal of this proposal is to investigate and build a thorough understanding of lifelong language learning (LLL). This project includes: (1) Extending LAMAL, our previous work on LLL, on the entire DecaNLP dataset. (2) Quantifying how semi-multitask guides LLL. (3) Obtaining a theoretical view of LLL in terms of neural network robustness and task relatedness. Eventually, we hope to realize real-world applications based on our research.

Pulkit Agrawal, Massachusetts Institute of Technology
Few-Shot Learning for Visual Recognition, Lifelong Learning, Multi-Task Learning
A Unified Framework for Continual and Few-Shot Learning

Current research has treated lifelong learning and transfer for few-shot recognition as separate problems. In lifelong learning, significant research effort is concerned with overcoming catastrophic forgetting by defining different ways to constrain weight updates. In transfer learning, meta-learning based methods have gained dominance in the past few years. Instead of treating these as separate problems, using the idea of weight superposition, we propose a unified framework that overcomes the problem of catastrophic forgetting allowing an agent to learn in a lifelong fashion by composing models learned for previous tasks. Because, solutions to new tasks are determined by composing previously learned models, we expect our system to outperform the current dominant paradigm of updating weights using finetuning in few-shot scenarios.

See Our 2018 Grant Winners Here

Application process

Applicants should submit a two-page PDF detailing their proposed research and the impact it will have on their respective research community. We encourage you to identify what your research would be enabling, and the proposed outcome should you achieve your vision. Please detail your proposed test case, expected challenges, and how you plan to mitigate them. Applicants should also include a budget overview of how the grant funding would be applied, and if you currently have additional funding for this area of research. Please provide links to previous research papers and citations. A third page can be used for references.

See our papers for an overview of recent research projects. Please mention any relevant papers and authors in your proposal. We have an incredible team of researchers at Salesforce who would love to partner with you on the grant.

Should you be selected we will notify you via email. Grant winners will be announced in our research blog and will be requested to provide an abstract that we will post to our website.

More information about the application and suggested research topics can be found in our FAQ.

10 AUG 2020
Grant announced and open to receive new applications.
16 OCT 2020
Application deadline by midnight pacific standard time.
Mid Nov 2020
Decisions notified by email.

Contact us

Feel free to send us any further questions.

Email us