HoT: Highlighted Chain of Thought for Referencing Supporting Facts from Inputs

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Nguyen, Tin, Bolton, Logan, Taesiri, Mohammad Reza, Bui, Trung, Nguyen, Anh Totti
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866912914037800960
author Nguyen, Tin
Bolton, Logan
Taesiri, Mohammad Reza
Bui, Trung
Nguyen, Anh Totti
author_facet Nguyen, Tin
Bolton, Logan
Taesiri, Mohammad Reza
Bui, Trung
Nguyen, Anh Totti
contents An Achilles heel of Large Language Models (LLMs) is their tendency to hallucinate non-factual statements. A response mixed of factual and non-factual statements poses a challenge for humans to verify and accurately base their decisions on. To combat this problem, we propose Highlighted Chain-of-Thought Prompting (HoT), a technique for prompting LLMs to generate responses with XML tags that ground facts to those provided in the question. That is, given an input question, LLMs would first re-format the question to add XML tags highlighting key facts, and then, generate a response with highlights over the facts referenced from the input. Compared to vanilla chain of thought prompting (CoT), HoT reduces the rate of hallucination and separately improves LLM accuracy consistently on over 22 tasks from arithmetic, reading comprehension, to logical reasoning. When asking humans to verify LLM responses, highlights help time-limited participants to more accurately and efficiently recognize when LLMs are correct. Yet, surprisingly, when LLMs are wrong, HoTs tend to fool users into believing that an answer is correct.
format Preprint
id arxiv_https___arxiv_org_abs_2503_02003
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle HoT: Highlighted Chain of Thought for Referencing Supporting Facts from Inputs
Nguyen, Tin
Bolton, Logan
Taesiri, Mohammad Reza
Bui, Trung
Nguyen, Anh Totti
Computation and Language
Human-Computer Interaction
An Achilles heel of Large Language Models (LLMs) is their tendency to hallucinate non-factual statements. A response mixed of factual and non-factual statements poses a challenge for humans to verify and accurately base their decisions on. To combat this problem, we propose Highlighted Chain-of-Thought Prompting (HoT), a technique for prompting LLMs to generate responses with XML tags that ground facts to those provided in the question. That is, given an input question, LLMs would first re-format the question to add XML tags highlighting key facts, and then, generate a response with highlights over the facts referenced from the input. Compared to vanilla chain of thought prompting (CoT), HoT reduces the rate of hallucination and separately improves LLM accuracy consistently on over 22 tasks from arithmetic, reading comprehension, to logical reasoning. When asking humans to verify LLM responses, highlights help time-limited participants to more accurately and efficiently recognize when LLMs are correct. Yet, surprisingly, when LLMs are wrong, HoTs tend to fool users into believing that an answer is correct.
title HoT: Highlighted Chain of Thought for Referencing Supporting Facts from Inputs
topic Computation and Language
Human-Computer Interaction
url https://arxiv.org/abs/2503.02003