# Evaluation of LLMs in Speech is Often Flawed: Test Set Contamination in Large Language Models for Speech Recognition

arXiv: https://arxiv.org/abs/2505.22251 · HTML: https://mlsanity.com/paper/2505.22251/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Builds on 33 of its 43 references that are in the map.
- First posted on arXiv 28 May 2025.

## Builds on (most cited first, 30 of 33)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [OPT: Open Pre-trained Transformer Language Models](https://mlsanity.com/paper/2205.01068/) (2022)
- [The Pile: An 800GB Dataset of Diverse Text for Language Modeling](https://mlsanity.com/paper/2101.00027/) (2020)
- [Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling](https://mlsanity.com/paper/2304.01373/) (2023)
- [WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processing](https://mlsanity.com/paper/2110.13900/) (2021)
- [Common Voice: A Massively-Multilingual Speech Corpus](https://mlsanity.com/paper/1912.06670/) (2019)
- [SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition](https://mlsanity.com/paper/1904.08779/) (2019)
- [The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networks](https://mlsanity.com/paper/1802.08232/) (2018)
- [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://mlsanity.com/paper/2204.06745/) (2022)
- [OLMo: Accelerating the Science of Language Models](https://mlsanity.com/paper/2402.00838/) (2024)
- [Dolma: an Open Corpus of Three Trillion Tokens for Language Model Pretraining Research](https://mlsanity.com/paper/2402.00159/) (2024)
- [SpeechBrain: A General-Purpose Speech Toolkit](https://mlsanity.com/paper/2106.04624/) (2021)
- [NLP Evaluation in trouble: On the Need to Measure LLM Data Contamination for each Benchmark](https://mlsanity.com/paper/2310.18018/) (2023)
- [Data Contamination: From Memorization to Exploitation](https://mlsanity.com/paper/2203.08242/) (2022)
- [Leak, Cheat, Repeat: Data Contamination and Evaluation Malpractices in Closed-Source LLMs](https://mlsanity.com/paper/2402.03927/) (2024)
- [RedPajama: an Open Dataset for Training Large Language Models](https://mlsanity.com/paper/2411.12372/) (2024)
- [The BigScience ROOTS Corpus: A 1.6TB Composite Multilingual Dataset](https://mlsanity.com/paper/2303.03915/) (2023)
- [On decoder-only architecture for speech-to-text and large language model integration](https://mlsanity.com/paper/2307.03917/) (2023)
- [Prompting Large Language Models with Speech Recognition Abilities](https://mlsanity.com/paper/2307.11795/) (2023)
- [WavLLM: Towards Robust and Adaptive Speech Large Language Model](https://mlsanity.com/paper/2404.00656/) (2024)
- [An Embarrassingly Simple Approach for LLM with Strong ASR Capacity](https://mlsanity.com/paper/2402.08846/) (2024)
- [Connecting Speech Encoder and Large Language Model for ASR](https://mlsanity.com/paper/2309.13963/) (2023)
- [Can Generative Large Language Models Perform ASR Error Correction?](https://mlsanity.com/paper/2307.04172/) (2023)
- [Open-Source Conversational AI with SpeechBrain 1.0](https://mlsanity.com/paper/2407.00463/) (2024)
- [COSMIC: Data Efficient Instruction-tuning For Speech In-Context Learning](https://mlsanity.com/paper/2311.02248/) (2023)
- [CTC-Assisted LLM-Based Contextual ASR](https://mlsanity.com/paper/2411.06437/) (2024)
- [Comparing Discrete and Continuous Space LLMs for Speech Recognition](https://mlsanity.com/paper/2409.00800/) (2024)
- [LLM360 K2: Building a 65B 360-Open-Source Large Language Model from Scratch](https://mlsanity.com/paper/2501.07124/) (2025)

## Built on by (most cited first, 9 of 9)

- [Open ASR Leaderboard: Towards Reproducible and Transparent Multilingual and Long-Form Speech Recognition Evaluation](https://mlsanity.com/paper/2510.06961/) (2025)
- [AQUA-Bench: Beyond Finding Answers to Knowing When There Are None in Audio Question Answering](https://mlsanity.com/paper/2601.12248/) (2026)
- [Walking Through Uncertainty: An Empirical Study of Uncertainty Estimation for Audio-Aware Large Language Models](https://mlsanity.com/paper/2604.25591/) (2026)
- [Data-Centric Lessons To Improve Speech-Language Pretraining](https://mlsanity.com/paper/2510.20860/) (2025)
- [Subtitle-Aligned Fine-Tuning of Whisper for Swiss German ASR: Benchmark Contamination, Convention Mismatch, and an Honest Baseline at 25.6% WER (13.8% cWER)](https://mlsanity.com/paper/2606.07608/) (2026)
- [Modeling Overlapped Speech with Shuffles](https://mlsanity.com/paper/2603.17769/) (2026)
- [Who Spoke What When? Evaluating Spoken Language Models for Conversational ASR with Semantic and Overlap-Aware Metrics](https://mlsanity.com/paper/2603.22709/) (2026)
- [Towards Quantifying Benchmark Optimization in ASR Models](https://mlsanity.com/paper/2608.19936/) (2026)
- [Personalized Korean Lipreading as Visual Speech Recognition: Transfer, Census and Adaptation on OLKAVS](https://mlsanity.com/paper/2609.28988/) (2026)

## Abstract (arXiv)

Recent work suggests that large language models (LLMs) can improve performance of speech tasks compared to existing systems. To support their claims, results on LibriSpeech and Common Voice are often quoted. However, this work finds that a substantial amount of the LibriSpeech and Common Voice evaluation sets appear in public LLM pretraining corpora. This calls into question the reliability of findings drawn from these two datasets. To measure contamination impact, LLMs trained with/without contamination are compared. A contaminated LLM is more likely to generate test sentences it has seen during training. Then, speech recognisers based on LLMs are compared. They show only subtle error rate differences if the LLM is contaminated, but assign significantly higher probabilities to transcriptions seen during LLM training. Results show that LLM outputs can be biased by tiny amounts of data contamination, highlighting the importance of evaluating LLM-based speech systems with held-out data.