Proposal

Identify a concrete problem you’re interested in solving for the project using an LLM.

Write the proposal in the proposal.qmd file in your project repo.

Proposal components

Problem or question

What is the problem you will solve?

• A well formulated objective.
• Statement on why this topic is important.
• Identify how LLMs will be used to solve the problem.

Data sources

You need to begin identifying potential datasets for your project as early as possible (assuming your project requires data).

• Identify the source of the data.

• State when and how it was originally collected (by the original data curator, not necessarily how you found the data).

• Write a brief description of the observations.

Final product

What will the final product of your project be? Explain what form your final product will take (e.g. interactive web application, analytical report, etc.).

Ethical concerns

Identify potential concerns about your use of an LLM. This could pertain to data privacy, cybersecurity, environmental impact, or other issues. Discuss how you will mitigate these concerns.

Project timeline

Provide a rough timeline for your project. Include a detailed timeline from proposal submission to final submission. Document all major activities that need to be completed in this time frame and when you will need to accomplish those tasks. Keep in mind you must have a draft of your product completed by December 4th for peer review.

Evaluation criteria

Category	Less developed projects	Typical projects	More developed projects
Problem formulation	Problem statement is vague or unclear. Role of LLMs in solving the problem is unclear or inappropriate.	Problem is clearly stated with reasonable scope. Explains why problem is important and how LLMs will be used to address it.	Problem is well-articulated with compelling motivation. Demonstrates deep understanding of problem space and provides innovative application of LLMs. Clear rationale for why LLMs are particularly suited to this problem.
Data sources	Data sources are not identified or are inappropriate for the project. No consideration of data availability or ethical concerns.	Identifies relevant data sources with basic information about origin and collection. Acknowledges data availability and mentions ethical considerations.	Provides comprehensive description of data sources including origin, collection methods, and observations. Demonstrates critical thinking about data ethics, biases, and limitations. Clear plan for data access and preprocessing.
Final product	Product format is unclear or inappropriate for the objectives. No consideration of user experience or accessibility.	Describes a reasonable product format appropriate for the problem. Indicates how product will be accessible to target audience.	Proposes an ambitious yet feasible product with clear value proposition. Thoughtful consideration of user experience, accessibility, and deployment. Product format aligns well with problem objectives and target users.
Ethical concerns	No discussion of ethical implications or concerns. No mitigation strategies proposed.	Identifies relevant ethical concerns related to LLM use. Proposes basic mitigation strategies for identified concerns.	Provides thorough analysis of multiple ethical dimensions. Proposes concrete, actionable mitigation strategies. Demonstrates awareness of broader societal implications and responsible AI practices.
Project timeline	Timeline is missing, vague, or unrealistic. Does not account for draft deadline or peer review requirements.	Provides reasonable timeline with major milestones identified. Timeline accounts for draft (Dec 4) and final submission deadlines.	Detailed, realistic timeline with well-defined milestones and dependencies. Timeline demonstrates careful planning with contingency for challenges. Clear allocation of work that will enable timely completion of all deliverables.
Feasibility	Scope is too limited or overly ambitious given timeline and resources. Unclear how team will complete project by deadline.	Project scope is reasonable and achievable within the semester. Clear path to completion with consideration of available resources and LLM access budget.	Project demonstrates appropriate ambition balanced with feasibility. Proposal anticipates potential challenges and includes strategies to address them. Demonstrates understanding of LLM capabilities and limitations relevant to proposed work.