The 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL) is being hosted in Albuquerque, New Mexico from April 29 - May 4. We’re excited to share all the work from SAIL that’s being presented, and you’ll find links to the papers below. Feel free to reach out to the contact authors directly to learn more about the work that’s happening at Stanford!
Main Conference
Benchmarking Distributional Alignment of Large Language Models
Contact: nmeist@stanford.edu
Links: Paper
Keywords: nlp tools for social analysis, language/cultural bias analysis, values and culture, benchmarking, nlp datasets, evaluation
Can Unconfident LLM Annotations Be Used for Confident Conclusions?
Contact: gligoric@stanford.edu
Award nominations: Best paper, outstanding paper
Links: Paper
Keywords: llm annotations, statistical inference, validity, nlp tools for social analysis, human behavior analysis
REL-A.I.: An Interaction-Centered Approach To Measuring Human-LM Reliance
Contact: katezhou@stanford.edu
Award nominations: Best Paper Runner Up
Links: Paper
Keywords: human-lm interaction, reliance, safety, expressions of uncertainty
Rethinking Word Similarity: Semantic Similarity through Classification Confusion
Contact: cshani@stanford.edu
Links: Paper
Keywords: semantic similarity, human-centered nlp, computational social science
Sneaking Syntax into Transformer Language Models with Tree Regularization
Contact: tgk.ananjan@gmail.com
Award nominations: Oral
Links: Paper | Website
Keywords: multi-task approaches, constituency parsing, syntactic language models, sample efficiency, syntactic generalization
“All that Glitters”: Techniques for Evaluations with Unreliable Model and Human Annotations
Contact: hardym@stanford.edu
Links: Paper | Website
Keywords: evaluation methodologies, model bias/fairness evaluation, model bias/unfairness mitigation, educational applications, ethical considerations in nlp applications, transparency, human-subject application-grounded evaluations, human-centered evaluation
Workshop Papers
Measuring Mental Health Variables in Computational Research: Toward Validated, Dimensional, and Transdiagnostic Approaches
Contact: cshani@stanford.edu
Workshop: CLPsych: The Workshop on Computational Linguistics and Clinical Psychology
Links: Paper
Keywords: nlp for cilnical psychology
What Can Large Language Models Do for Sustainable Food?
Contact: gligoric@stanford.edu
Workshop: First Workshop on AI and Scientific Discovery: Directions and Opportunities (AISD)
Links: Paper
Keywords: large language models, sustainability, climate, food, health, optimization
We look forward to seeing you at NAACL 2025!