SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2404.01651 · NAACL · 2024

NLP Systems That Can't Tell Use from Mention Censor Counterspeech, but Teaching the Distinction Helps

Dan Jurafsky, Myra Cheng, Esin Durmus, Lucia Zheng, Kristina Gligorić

arXiv · PDF · Open in the Atlas

Code that ran

We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

The use of words to convey speaker's intent is traditionally distinguished from the 'mention' of words for quoting what someone said, or pointing out properties of a word. Here we show that computationally modeling this usemention distinction is crucial for dealing with counterspeech online. Counterspeech that refutes problematic content often mentions harmful language but is not harmful itself (e.g., calling a vaccine dangerous is not the same as expressing disapproval of someone for calling vaccines dangerous). We show that even recent language models fail at distinguishing use from mention, and that this failure propagates to two key downstream tasks: misinformation and hate speech detection, resulting in censorship of counterspeech. We introduce prompting mitigations that teach the use-mention distinction, and show they reduce these errors. Our work highlights the importance of the use-mention distinction for NLP and CSS and offers ways to address it.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2404.01651")
get_code_for_paper("2404.01651")
have("2404.01651")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2404.01651.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2404.01651.svg)](https://syntology.ai/paper/2404.01651)

Connect an agent — have() is free.