# Preference-ASR: A Preference-Aware Test Set for Benchmarking ASR in the Era of Speech LLMs

arXiv: https://arxiv.org/abs/2606.29534 · HTML: https://mlsanity.com/paper/2606.29534/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 25 of its 33 references that are in the map.
- First posted on arXiv 28 Jun 2026.

## Builds on (most cited first, 25 of 25)

- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [Robust Speech Recognition via Large-Scale Weak Supervision](https://mlsanity.com/paper/2212.04356/) (2022)
- [Common Voice: A Massively-Multilingual Speech Corpus](https://mlsanity.com/paper/1912.06670/) (2019)
- [Qwen2.5-Omni Technical Report](https://mlsanity.com/paper/2503.20215/) (2025)
- [Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models](https://mlsanity.com/paper/2311.07919/) (2023)
- [Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAs](https://mlsanity.com/paper/2503.01743/) (2025)
- [SALMONN: Towards Generic Hearing Abilities for Large Language Models](https://mlsanity.com/paper/2310.13289/) (2023)
- [Qwen3-Omni Technical Report](https://mlsanity.com/paper/2509.17765/) (2025)
- [VoxPopuli: A Large-Scale Multilingual Speech Corpus for Representation Learning, Semi-Supervised Learning and Interpretation](https://mlsanity.com/paper/2101.00390/) (2021)
- [GigaSpeech: An Evolving, Multi-domain ASR Corpus with 10,000 Hours of Transcribed Audio](https://mlsanity.com/paper/2106.06909/) (2021)
- [Deep context: end-to-end contextual speech recognition](https://mlsanity.com/paper/1808.02480/) (2018)
- [Fast Conformer with Linearly Scalable Attention for Efficient Speech Recognition](https://mlsanity.com/paper/2305.05084/) (2023)
- [SPGISpeech: 5,000 hours of transcribed financial audio for fully formatted end-to-end speech recognition](https://mlsanity.com/paper/2104.02014/) (2021)
- [Canary-1B-v2 & Parakeet-TDT-0.6B-v3: Efficient and High-Performance Models for Multilingual ASR and AST](https://mlsanity.com/paper/2509.14128/) (2025)
- [Efficient Sequence Transduction by Jointly Predicting Tokens and Durations](https://mlsanity.com/paper/2304.06795/) (2023)
- [Earnings-22: A Practical Benchmark for Accents in the Wild](https://mlsanity.com/paper/2203.15591/) (2022)
- [RNN Approaches to Text Normalization: A Challenge](https://mlsanity.com/paper/1611.00068/) (2016)
- [Speech-IFEval: Evaluating Instruction-Following and Quantifying Catastrophic Forgetting in Speech-Aware Language Models](https://mlsanity.com/paper/2505.19037/) (2025)
- [MMAU: A Holistic Benchmark of Agent Capabilities Across Diverse Domains](https://mlsanity.com/paper/2407.18961/) (2024)
- [What is lost in Normalization? Exploring Pitfalls in Multilingual ASR Model Evaluations](https://mlsanity.com/paper/2409.02449/) (2024)
- [Instruction-Following Speech Recognition](https://mlsanity.com/paper/2309.09843/) (2023)
- [Longer is (Not Necessarily) Stronger: Punctuated Long-Sequence Training for Enhanced Speech Recognition and Translation](https://mlsanity.com/paper/2409.05601/) (2024)
- [End-to-End Rich Transcription-Style Automatic Speech Recognition with Semi-Supervised Learning](https://mlsanity.com/paper/2107.05382/) (2021)

## Built on by (most cited first, 1 of 1)

- [Voice Memory for Agentic Speech Recognition](https://mlsanity.com/paper/2607.26410/) (2026)

## Abstract (arXiv)

Popular ASR test sets adopt inconsistent conventions for numbers, disfluencies, entities, and casing, while standard normalizers erase the format distinctions users care about. Current benchmarks therefore cannot measure whether a model follows user preferences for output style. We introduce PreferenceASR, a test set evaluating ASR systems on their ability to follow natural-language preference instructions across four categories: normalization, entities, disfluencies, and case. Built from seven open-source corpora via a two-stage LLM-assisted pipeline with human verification, it is evaluated with a preference-aware normalizer that selectively skips steps matching the active instruction. Benchmarking four models shows rankings shift across preference types, exposing quality differences traditional evaluation obscures. We publicly release the dataset.