Cathy Jiao

Email: cljiao@cs.cmu.edu

I am a PhD student at the Language Technologies Institute in the School of Computer Science at Carnegie Mellon University, advised by Chenyan Xiong.

My research focuses on data-centric AI. I aim to develop frameworks that make data usage more transparent, reliable, and impactful for both research and deployment of foundation models. This includes empirical analyses of the accuracy and cost of measuring data value for LLMs, benchmarks for evaluating such methods on real-world tasks, and applications such as data pricing.

Previously, I finished my MS at CMU LTI where I worked on dialogue systems, advised by Maxine Eskenazi and Aaron Steinfeld. Prior to that, I graduated with distinction from the University of British Columbia with a B.S. in CS & Math.

News

Aug 13, 2026	Gave a talk at the Jane Street Research Symposium on synthetic data curation for LLMs.
Apr 24, 2026	Work from my Spotify internship (dataset curation for generative recommenders), was presented at the ICLR 2026 CAO Workshop (top 8% of accepted papers). [slides].
Sep 18, 2025	DATE-LM was accepted to NeurIPS 2025. We introduce a rigorous, applications-driven benchmark for large-scale evaluation of data attribution methods in LLMs.
Sep 18, 2025	Fairshare Data Pricing was accepted to NeurIPS 2025, introducing a data-influence–based framework for fair pricing of LLM training datasets.
Aug 15, 2025	Gave a talk at Spotify on ICP for Data Attribution [slides].
Feb 01, 2025	ICP for Data Attribution was accepted to NAACL 2025, showing that probing LLMs is a cheap proxy for gradient-based attribution of influential training samples.

Selected Publications

(See Google Scholar for all)

*= equal contribution

Efficient Dataset Selection for Continual Adaptation of Generative Recommenders

Cathy Jiao , Juan Elenter , Praveen Ravichandran , Bernd Huber , Joseph Cauteruccio , Todd Wasson , Timothy Heath , Chenyan Xiong , Mounia Lalmas , and Paul Bennett

In ICLR CAO Workshop, 2026

Oral Presentation (top 8%) Abs PDF Slides

Selected for an oral presentation at the ICLR 2026 CAO Workshop (6 of 76 papers, top 8%).

Recommendation systems must continuously adapt to evolving user behavior, yet the volume of data generated in large-scale streaming environments makes frequent full retraining impractical. This work investigates how targeted data selection can mitigate performance degradation caused by temporal distributional drift while maintaining scalability. We evaluate a range of representation choices and sampling strategies for curating small but informative subsets of user interaction data. Our results demonstrate that gradient-based representations, coupled with distribution-matching, improve downstream model performance, achieving training efficiency gains while preserving robustness to drift. These findings highlight data curation as a practical mechanism for scalable monitoring and adaptive model updates in production-scale recommendation systems.
DATE-LM: Benchmarking Data Attribution Evaluation for Large Language Models

Cathy Jiao^* , Yijun Pan^* , Emily Xiao^* , Daisy Sheng , Niket Jain , Hanzhang Zhao , Ishita Dasgupta , Jiaqi W. Ma , and Chenyan Xiong

NeurIPS, 2025

Abs PDF Code

Data attribution methods quantify the influence of training data on model outputs and are becoming increasingly relevant for a wide range of LLM research and applications, including dataset curation, model interpretability, data valuation. However, there remain critical gaps in systematic LLM-centric evaluation of data attribution methods. To this end, we introduce DATE-LM (Data Attribution Evaluation in Language Models), a unified benchmark for evaluating data attribution methods through real-world LLM applications. DATE-LM measures attribution quality through three key tasks – training data selection, toxicity/bias filtering, and factual attribution. Our benchmark is designed for ease of use, enabling researchers to configure and run large-scale evaluations across diverse tasks and LLM architectures. Furthermore, we use DATE-LM to conduct a large-scale evaluation of existing data attribution methods. Our findings show that no single method dominates across all tasks, data attribution methods have trade-offs with simpler baselines, and method performance is sensitive to task-specific evaluation design. Finally, we release a public leaderboard for quick comparison of methods and to facilitate community engagement. We hope DATE-LM serves as a foundation for future data attribution research in LLMs.
Fairshare Data Pricing via Data Valuation for Large Language Models

Luyang Zhang^* , Cathy Jiao^* , Beibei Li , and Chenyan Xiong

NeurIPS, 2025

Abs PDF

Training data is a pivotal resource for building large language models (LLMs), but unfair pricing in data markets poses a serious challenge for both data buyers (e.g., LLM builders) and sellers (e.g., human annotators), which discourages market participation, reducing data quantity and quality. In this paper, we propose a fairshare pricing framework that sets training data prices using data valuation methods to quantify their contribution to LLMs. In our framework, buyers make purchasing decisions using data valuation and sellers set prices to maximize their profits based on the anticipated buyer purchases. We theoretically show that pricing derived from our framework is tightly linked to data valuation and buyers’ budget, optimal for both buyers and sellers. Through market simulations using current LLMs and datasets (math problems, medical diagnosis, and physical reasoning), we show that our framework is fairshare for buyers by ensuring their purchased data is reflective of model training value, leading to higher LLM task performances per-dollar spent on data, and fairshare for sellers by ensuring they sell their data at optimal prices. Our framework lays the foundation for future research on equitable and sustainable data markets for large-scale AI.
On the Feasibility of In-Context Probing for Data Attribution

Cathy Jiao , Gary Gao , Aditi Raghunathan , and Chenyan Xiong

NAACL, 2025

Abs PDF Code

Data attribution methods are used to measure the contribution of training data towards model outputs, and have several important applications in areas such as dataset curation and model interpretability. However, many standard data attribution methods, such as influence functions, utilize model gradients and are computationally expensive. In our paper, we show in-context probing (ICP) – prompting a LLM – can serve as a fast proxy for gradient-based data attribution for data selection under conditions contingent on data similarity. We study this connection empirically on standard NLP tasks, and show that ICP and gradient-based data attribution are well-correlated in identifying influential training data for tasks that share similar task type and content as the training data. Additionally, fine-tuning models on influential data selected by both methods achieves comparable downstream performance, further emphasizing their similarities. We also examine the connection between ICP and gradient-based data attribution using synthetic data on linear regression tasks. Our synthetic data experiments show similar results with those from NLP tasks, suggesting that this connection can be isolated in simpler settings, which offers a pathway to bridging their differences.
Improving Zero and Few-shot Generalization in Dialogue through Instruction Tuning

Prakhar Gupta , Cathy Jiao , Yi-Ting Yeh , Shikib Mehri , Maxine Eskenazi , and Jeffrey P. Bigham

EMNLP, 2022

Abs DOI PDF Code

Instruction tuning is an emergent paradigm in NLP wherein natural language instructions are leveraged with language models to induce zero-shot performance on unseen tasks. Instructions have been shown to enable good performance on unseen tasks and datasets in both large and small language models. Dialogue is an especially interesting area to explore instruction tuning because dialogue systems perform multiple kinds of tasks related to language (e.g., natural language understanding and generation, domain-specific interaction), yet instruction tuning has not been systematically explored for dialogue-related tasks. We introduce InstructDial, an instruction tuning framework for dialogue, which consists of a repository of 48 diverse dialogue tasks in a unified text-to-text format created from 59 openly available dialogue datasets. Next, we explore cross-task generalization ability on models tuned on InstructDial across diverse dialogue tasks. Our analysis reveals that InstructDial enables good zero-shot performance on unseen datasets and tasks such as dialogue evaluation and intent detection, and even better performance in a few-shot setting. To ensure that models adhere to instructions, we introduce novel meta-tasks. We establish benchmark zero-shot and few-shot performance of models trained using the proposed framework on multiple dialogue tasks.