Afra Amini

In-Context Probing: Toward Building Robust Classifiers via Probing Large Language Models

2023 · arXiv (Cornell University)

Large language models are able to learn new tasks in context, where they are provided with instructions and a few annotated examples. However, the effectiveness of in-context learning is dependent on the provided context, and …

Naturalistic Causal Probing for Morpho-Syntax

2023 · Transactions of the Association for Computational Linguistics

Abstract Probing has become a go-to methodology for interpreting and analyzing deep neural models in natural language processing. However, there is still a lack of understanding of the limitations and weaknesses of various types of …

Variational Best-of-N Alignment

2024 · arXiv (Cornell University)

Best-of-N (BoN) is a popular and effective algorithm for aligning language models to human preferences. The algorithm works as follows: at inference time, N samples are drawn from the language model, and the sample with …