2.3k citations · 2.9k across the 25 of their papers we have counts for
26 papers
Generative Classifiers Avoid Shortcut Solutions
Alexander C. Li, Ananya Kumar, Deepak Pathak
Discriminative approaches to classification often learn shortcuts that hold in-distribution but fail even under minor distribution shift. This failure mode stems from an overrelian…
Language Models Prefer What They Know: Relative Confidence Estimation via Confidence Preferences
Vaishnavi Shrivastava, Ananya Kumar, Percy Liang
Language models (LMs) should provide reliable confidence estimates to help users detect mistakes in their outputs and defer to human experts when necessary. Asking a language model…
Evolving Domain Adaptation of Pretrained Language Models for Text Classification
Yun-Shiuan Chuang, Yi Wu, Dhruv Gupta +7
Adapting pre-trained language models (PLMs) for time-series text classification amidst evolving domain shifts (EDS) is critical for maintaining accuracy in applications like stance…
Llamas Know What GPTs Don't Show: Surrogate Models for Confidence Estimation
Vaishnavi Shrivastava, Percy Liang, Ananya Kumar
To maintain user trust, large language models (LLMs) should signal low confidence on examples where they are incorrect, instead of misleading the user. The standard approach of est…
Conservative Prediction via Data-Driven Confidence Minimization
Caroline Choi, Fahim Tajwar, Yoonho Lee +3
In safety-critical applications of machine learning, it is often desirable for a model to be conservative, abstaining from making predictions on unknown inputs which are not well-r…
Improving Representational Continuity via Continued Pretraining
Michael Sun, Ananya Kumar, Divyam Madaan +1
We consider the continual representation learning setting: sequentially pretrain a model on tasks , and then adapt on a small amount of data from each t…