At Cerebras, we believe AI is the most transformative technology of our generation. Our mission is to accelerate AI by making it faster, easier to use, and more energy efficient, making AI accessible around the world. 

Cerebras inference, powered by the third generation Wafer Scale Engine, is the fastest AI inference solution in the world. Cerebras inference delivers over 2,100 tokens per second for Llama-3.1-70b, 16x faster than the fastest GPU solution. 

We invite university faculty and researchers to respond to this Request for Proposals (RFP) to advance the field of Generative AI. We aim to support innovative research that will leverage our technology to drive forward new techniques, applications, and insights in AI. 

Through this RFP, we seek collaborations that will push the boundaries of state-of-the-art AI research and contribute to the broader scientific and technological communities. 

Grant 

Selected Principal Investigators (PIs) may receive the following: 

• • Cerebras Inference service credits, equivalent to no more than $50,000 USD of inference credits 
  • Cerebras tutorials and hands-on sessions with Cerebras engineers and researchers 

The final grant amount and details will be determined by Cerebras. 

Areas of Interest 

Proposals should focus on the area of generative AI, leveraging Cerebras Inference and the inference token grants.  

We recommend that proposals are attached to one or more of the areas listed below, although we will consider any proposal that leverages inference tokens to advance generative AI. 

Recommended proposal areas: 

• • How to improve inference quality using inference-time compute.
    Areas of research may include:
    •  Prompting techniques such as chain-of-thought, ReACT, programmable prompt customization, etc. 
    • In-Context Learning (ICL).
    • Inference decoding algorithms such as sampling, chain-of-thought decoding, contrastive decoding, etc. 
    • Meta-generation algorithms that incorporate the language model within larger generation programs, such as best-of-N, majority voting, self-consistency, tool-calling, etc. 
    • Leveraging additional knowledge base through a combination of iterative multi-step retrieval, reasoning, and generation. 

• • How to enable long form generation.
    Areas of research may include: 

• • • Planning content structure, content style, analogies etc. 
    • Iterative writing, reflection, and refinement. 
    • Editing document based on user guidance to collaboratively produce the content. 

• • How to perform model evaluations, quality benchmarks, and safety benchmarks.
    Areas of research may include: 

• • • Development of robust benchmarks and metrics to assess model quality, reliability, safety, and bias.  
    • Novel approaches to operate generative models responsibly, focusing on measuring, preventing, and mitigating hallucinations, biases, and other harmful or undesirable outputs. 

• • How to generate high-quality synthetic data.
    Areas of research may include: 

• • • Novel methods to generate large-scale synthetic datasets with precisely defined properties using generative models. 
    • Filtering techniques to ensure quality and diversity of the generated samples

Timeline 

Submission period: Dec 10 2024 to Jan 31 2025 11:59pm PT. 

Decision letters will be sent out February 2025 

Application Requirements 

Prior to submitting a proposal, please confirm that your institution will agree to a contract for sponsored research from Cerebras. You may assume that all IP rights shall belong to the university and the agreement will operate under “Open Science” terms meaning that all research will be available for publication in the public sphere.  

Application materials should include: 

• • An English summary of the project (1–2 pages) explaining the area of focus, a description of techniques, how inference credits will be used, and a timeline with milestones and expected outcomes. 
  • Contact details of the Principal Investigator: 
    • Full name 
    • Title / Position 
    • Academic institute 
    • Department 
    • Email address 
    • Phone number 
    • Mailing address 
    • Country 
    • Curriculum Vitae for all project participants. 

Eligibility 

Proposals must comply with applicable US and international laws, regulations and policies. 

Applicants must be current full-time faculty at an accredited academic institution that awards research degrees to PhD students. 

Applicants must be the Principal Investigator on any resulting grant. 

Cerebras cannot consider proposals submitted, prepared or to be carried out by individuals residing in, or affiliated with an academic institution located in, a country or territory subject to comprehensive U.S. trade sanctions or is otherwise not permitted to access Cerebras inference. 

Government officials (excluding faculty and staff of public universities, to the extent they may be considered government officials), political figures, and politically affiliated businesses (all as determined by Cerebras in its sole discretion) are not eligible. 

Terms and Conditions 

Please read these terms carefully before proceeding. 

• • Cerebras decisions will be final in all matters relating to RFP solicitations, including whether or not to select a proposal for funding and the interpretation of Cerebras RFP Terms and Conditions. By submitting a proposal, applicants affirm that they have read and agree to these Terms and Conditions. 
  • Cerebras is authorized to evaluate proposals submitted under its RFPs and to consult with outside experts as needed. Applicants agree not to challenge any such decisions.  
  • Cerebras will not be required to treat any part of a proposal as confidential or protected by copyright, and may use, edit, modify, copy, reproduce and distribute all or a portion of the proposal in any manner for the sole purposes of administering the RFP and evaluating the contents of the proposal. 
  • Personal data submitted with a proposal, including name, mailing address, phone number, and email address of the applicant and other named researchers in the proposal may be collected, processed, stored and otherwise used by Cerebras for the purposes of evaluating the contents of the proposal.  
  • Cerebras is under no obligation to review or consider the proposal. 
  • All credits or funding provided by Cerebras in connection with this RFP shall be used only in accordance with applicable laws and shall not be used in any way, directly or indirectly, to facilitate any act that would constitute bribery or an illegal kickback, an illegal campaign contribution, or would otherwise violate any applicable anti-corruption, or  political activities, export or sanctions laws.  
  • Applicants understand and acknowledge that their organization will need to agree to the terms and conditions of such sponsored research agreement to receive funding for the RFP proposal.