
Greg Durrett
Associate Professor of Computer Science and Data Science · Department of Computer Science and Center for Data Science · New York University
At Prof Rod’s reading desk
Questions to explore
- What evidence is enough to support a generated claim?
- How can a smaller checker be compared fairly with a larger language model?
Research interests
Greg Durrett works on natural language processing, language-model reasoning and factuality evaluation.
- Natural language processing
- Language-model reasoning
- Factuality evaluation
Explore the work
Prof Rod’s reading notes
Start with a paper
EMNLP · 2024
MiniCheck: Efficient Fact-Checking of LLMs on Grounding Documents
Prof Rod’s reading note: Study how smaller models check claims against source documents. Compare accuracy and cost under the paper’s evaluation conditions.
Bring a question of your own
Turn reading into research
Choose one generated claim and its source passage. Predict whether the passage supports the whole claim, then identify a small edit that would make the claim unsupported.
Keep exploring
More at New York University
Explore the universityImprove this page
Suggest a correction, share a paper to start with, or point readers to a group page or opening.