Emotional RAG LLMs: Reading Comprehension for the Open Internet

TL;DR

This paper addresses the challenge of interpreting emotionally diverse internet texts in retrieval-augmented LLMs by creating a new dataset, an emotion translation model, and a prompt-based approach to enhance understanding.

Contribution

It introduces a novel dataset of emotionally inflected texts, an emotion translation model, and a prompt-based method to improve LLMs' pragmatic comprehension of diverse internet-based content.

Findings

Enhanced LLM understanding of emotional and sarcastic texts

Improved performance with emotion-aware prompts

Demonstrated effectiveness on real-world internet data

Abstract

Queries to large language models (LLMs) can be divided into two parts: the instruction/question and the accompanying context. The context for retrieval-augmented generation (RAG) systems in most benchmarks comes from Wikipedia-like texts written in a neutral and factual tone. However, real-world RAG applications often retrieve internet-based text with diverse tones and linguistic styles, posing challenges for downstream tasks. This paper introduces (a) a dataset that transforms RAG-retrieved passages into emotionally inflected and sarcastic text, (b) an emotion translation model for adapting text to different tones, and (c) a prompt-based method to improve LLMs' pragmatic interpretation of retrieved text.