# Enhancing Naturalness in LLM-Generated Utterances through Disfluency Insertion

arXiv: https://arxiv.org/abs/2412.12710 · HTML: https://mlsanity.com/paper/2412.12710/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 9 of its 33 references that are in the map.
- First posted on arXiv 17 Dec 2024.

## Builds on (most cited first, 9 of 9)

- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [BERTScore: Evaluating Text Generation with BERT](https://mlsanity.com/paper/1904.09675/) (2019)
- [SEP-28k: A Dataset for Stuttering Event Detection From Podcasts With People Who Stutter](https://mlsanity.com/paper/2102.12394/) (2021)
- [Disfluencies and Human Speech Transcription Errors](https://mlsanity.com/paper/1904.04398/) (2019)
- [Disfluency Detection with Unlabeled Data and Small BERT Models](https://mlsanity.com/paper/2104.10769/) (2021)
- [LARD: Large-scale Artificial Disfluency Generation](https://mlsanity.com/paper/2201.05041/) (2022)
- [A General-purpose AI Avatar in Healthcare](https://mlsanity.com/paper/2401.12981/) (2024)

## Built on by (most cited first, 3 of 3)

- [TriageSim: A Conversational Emergency Triage Simulation Framework from Structured Electronic Health Records](https://mlsanity.com/paper/2603.10035/) (2026)
- [Why Synthetic Isn't Real Yet: A Diagnostic Framework for Contact Center Dialogue Generation](https://mlsanity.com/paper/2508.18210/) (2025)
- [The Role of Disfluencies in Speech Translation](https://mlsanity.com/paper/2608.02138/) (2026)

## Abstract (arXiv)

Disfluencies are a natural feature of spontaneous human speech but are typically absent from the outputs of Large Language Models (LLMs). This absence can diminish the perceived naturalness of synthesized speech, which is an important criteria when building conversational agents that aim to mimick human behaviours. We show how the insertion of disfluencies can alleviate this shortcoming. The proposed approach involves (1) fine-tuning an LLM with Low-Rank Adaptation (LoRA) to incorporate various types of disfluencies into LLM-generated utterances and (2) synthesizing those utterances using a text-to-speech model that supports the generation of speech phenomena such as disfluencies. We evaluated the quality of the generated speech across two metrics: intelligibility and perceived spontaneity. We demonstrate through a user study that the insertion of disfluencies significantly increase the perceived spontaneity of the generated speech. This increase came, however, along with a slight reduction in intelligibility.