For Those Who May Find Themselves on the Red Team

TL;DR

This paper advocates for literary scholars to engage with LLM interpretability research, emphasizing the importance of critical involvement in red team activities to challenge current interpretability standards.

Contribution

It highlights the need for literary scholars to participate in LLM interpretability and red team efforts, bridging humanities and AI research.

Findings

Engagement with LLM interpretability is crucial for literary scholars.

Red team activities can serve as a site for ideological and interpretive struggle.

Current interpretability approaches may be insufficient for literary analysis.

Abstract

This position paper argues that literary scholars must engage with large language model (LLM) interpretability research. While doing so will involve ideological struggle, if not out-right complicity, the necessity of this engagement is clear: the abiding instrumentality of current approaches to interpretability cannot be the only standard by which we measure interpretation with LLMs. One site at which this struggle could take place, I suggest, is the red team.