# Thinking with Sound: Audio Chain-of-Thought Enables Multimodal Reasoning in Large Audio-Language Models

arXiv: https://arxiv.org/abs/2509.21749 · HTML: https://mlsanity.com/paper/2509.21749/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Placed in Audio-language model reasoning, one of 17 directions in LLM reasoning.
- Builds on 25 of its 25 references that are in the map.
- First posted on arXiv 26 Sep 2025.

## Placed in

- LLM reasoning / [Audio-language model reasoning](https://mlsanity.com/lr/audio-language-model-reasoning/)

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

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Large Language Models are Zero-Shot Reasoners](https://mlsanity.com/paper/2205.11916/) (2022)
- [Toolformer: Language Models Can Teach Themselves to Use Tools](https://mlsanity.com/paper/2302.04761/) (2023)
- [Robust Speech Recognition via Large-Scale Weak Supervision](https://mlsanity.com/paper/2212.04356/) (2022)
- [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://mlsanity.com/paper/2006.11477/) (2020)
- [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://mlsanity.com/paper/2106.07447/) (2021)
- [HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face](https://mlsanity.com/paper/2303.17580/) (2023)
- [Multimodal Chain-of-Thought Reasoning in Language Models](https://mlsanity.com/paper/2302.00923/) (2023)
- [Qwen2.5-Omni Technical Report](https://mlsanity.com/paper/2503.20215/) (2025)
- [MELD: A Multimodal Multi-Party Dataset for Emotion Recognition in Conversations](https://mlsanity.com/paper/1810.02508/) (2018)
- [Moshi: a speech-text foundation model for real-time dialogue](https://mlsanity.com/paper/2410.00037/) (2024)
- [Qwen2-Audio Technical Report](https://mlsanity.com/paper/2407.10759/) (2024)
- [SALMONN: Towards Generic Hearing Abilities for Large Language Models](https://mlsanity.com/paper/2310.13289/) (2023)
- [AudioGPT: Understanding and Generating Speech, Music, Sound, and Talking Head](https://mlsanity.com/paper/2304.12995/) (2023)
- [AIR-Bench: Benchmarking Large Audio-Language Models via Generative Comprehension](https://mlsanity.com/paper/2402.07729/) (2024)
- [LLaMA-Omni: Seamless Speech Interaction with Large Language Models](https://mlsanity.com/paper/2409.06666/) (2024)
- [GAMA: A Large Audio-Language Model with Advanced Audio Understanding and Complex Reasoning Abilities](https://mlsanity.com/paper/2406.11768/) (2024)
- [AudioBench: A Universal Benchmark for Audio Large Language Models](https://mlsanity.com/paper/2406.16020/) (2024)
- [Interleaved-Modal Chain-of-Thought](https://mlsanity.com/paper/2411.19488/) (2024)
- [Salmon: A Suite for Acoustic Language Model Evaluation](https://mlsanity.com/paper/2409.07437/) (2024)
- [EmphAssess : a Prosodic Benchmark on Assessing Emphasis Transfer in Speech-to-Speech Models](https://mlsanity.com/paper/2312.14069/) (2023)
- [MusicAgent: An AI Agent for Music Understanding and Generation with Large Language Models](https://mlsanity.com/paper/2310.11954/) (2023)
- [AHELM: A Holistic Evaluation of Audio-Language Models](https://mlsanity.com/paper/2508.21376/) (2025)
- [Towards Reliable Large Audio Language Model](https://mlsanity.com/paper/2505.19294/) (2025)

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

- [CamReasoner: Reinforcing Camera Movement Understanding via Structured Spatial Reasoning](https://mlsanity.com/paper/2602.00181/) (2026)
- [LatentOmni: Rethinking Omni-Modal Understanding via Unified Audio-Visual Latent Reasoning](https://mlsanity.com/paper/2605.22012/) (2026)
- [VideoStir: Understanding Long Videos via Spatio-Temporally Structured and Intent-Aware RAG](https://mlsanity.com/paper/2604.05418/) (2026)
- [Audio Interaction Model](https://mlsanity.com/paper/2606.05121/) (2026)
- [SEE: Signal Embedding Energy for Quantifying Noise Interference in Large Audio Language Models](https://mlsanity.com/paper/2601.07331/) (2026)
- [Entity Binding Failures in Speech LLM Reasoning: Diagnosis and Chain-of-Thought Intervention](https://mlsanity.com/paper/2606.04474/) (2026)
- [CogSR: Semantic-Aware Speech Super-Resolution via Chain-of-Thought Guided Flow Matching](https://mlsanity.com/paper/2512.16304/) (2025)
- [Do Audio LLMs Listen Before They Act? Diagnosing Acoustic-Context Gating in Voice Agents](https://mlsanity.com/paper/2609.32536/) (2026)
- [Long-Term Memory-Guided Enhancement for Target Perception in Audio-Language Models](https://mlsanity.com/paper/2609.36577/) (2026)

## Abstract (arXiv)

Recent Large Audio-Language Models (LALMs) have shown strong performance on various audio understanding tasks such as speech translation and Audio Q&A. However, they exhibit significant limitations on challenging audio reasoning tasks in complex acoustic scenarios. These situations would greatly benefit from the use of acoustic tools like noise suppression, source separation, and precise temporal alignment, but current LALMs lack access to such tools. To address this limitation, we introduce Thinking-with-Sound (TwS), a framework that equips LALMs with Audio CoT by combining linguistic reasoning with on-the-fly audio-domain analysis. Unlike existing approaches that treat audio as static input, TwS enables models to actively think with audio signals, performing numerical analysis and digital manipulation through multimodal reasoning. To evaluate this approach, we construct MELD-Hard1k, a new robustness benchmark created by introducing various acoustic perturbations. Experiments reveal that state-of-the-art LALMs suffer dramatic performance degradation on MELD-Hard1k, with accuracy dropping by more than 50% compared to clean audio. TwS achieves substantial improvements in robustness, demonstrating both effectiveness and scalability: small models gain 24.73% absolute accuracy, with improvements scaling consistently up to 36.61% for larger models. Our findings demonstrate that Audio CoT can significantly enhance robustness without retraining, opening new directions for developing more robust audio understanding systems.