# Audio-DeepThinker: Progressive Reasoning-Aware Reinforcement Learning for High-Quality Chain-of-Thought Emergence in Audio Language Models

arXiv: https://arxiv.org/abs/2604.18187 · HTML: https://mlsanity.com/paper/2604.18187/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Audio-language model reasoning, one of 17 directions in LLM reasoning.
- Builds on 24 of its 42 references that are in the map.
- First posted on arXiv 20 Apr 2026.

## Placed in

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

## Builds on (most cited first, 24 of 24)

- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [GPT-4o System Card](https://mlsanity.com/paper/2410.21276/) (2024)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism](https://mlsanity.com/paper/1909.08053/) (2019)
- [OpenAI o1 System Card](https://mlsanity.com/paper/2412.16720/) (2024)
- [DeepSeek-V3 Technical Report](https://mlsanity.com/paper/2412.19437/) (2024)
- [Qwen2.5-Omni Technical Report](https://mlsanity.com/paper/2503.20215/) (2025)
- [Vision-R1: Incentivizing Reasoning Capability in Multimodal Large Language Models](https://mlsanity.com/paper/2503.06749/) (2025)
- [Qwen2-Audio Technical Report](https://mlsanity.com/paper/2407.10759/) (2024)
- [Qwen3-Omni Technical Report](https://mlsanity.com/paper/2509.17765/) (2025)
- [Kimi-Audio Technical Report](https://mlsanity.com/paper/2504.18425/) (2025)
- [Audio Flamingo 3: Advancing Audio Intelligence with Fully Open Large Audio Language Models](https://mlsanity.com/paper/2507.08128/) (2025)
- [GDPO: Group reward-Decoupled Normalization Policy Optimization for Multi-reward RL Optimization](https://mlsanity.com/paper/2601.05242/) (2026)
- [Baichuan-Omni-1.5 Technical Report](https://mlsanity.com/paper/2501.15368/) (2025)
- [Step-Audio 2 Technical Report](https://mlsanity.com/paper/2507.16632/) (2025)
- [Reinforcement Learning Outperforms Supervised Fine-Tuning: A Case Study on Audio Question Answering](https://mlsanity.com/paper/2503.11197/) (2025)
- [Omni-R1: Do You Really Need Audio to Fine-Tune Your Audio LLM?](https://mlsanity.com/paper/2505.09439/) (2025)
- [Audio-Thinker: Guiding Audio Language Model When and How to Think via Reinforcement Learning](https://mlsanity.com/paper/2508.08039/) (2025)
- [Measuring Audio's Impact on Correctness: Audio-Contribution-Aware Post-Training of Large Audio Language Models](https://mlsanity.com/paper/2509.21060/) (2025)
- [Incentivizing Consistent, Effective and Scalable Reasoning Capability in Audio LLMs via Reasoning Process Rewards](https://mlsanity.com/paper/2510.20867/) (2025)
- [Covo-Audio Technical Report](https://mlsanity.com/paper/2602.09823/) (2026)
- [The Interspeech 2026 Audio Reasoning Challenge: Evaluating Reasoning Process Quality for Audio Reasoning Models and Agents](https://mlsanity.com/paper/2602.14224/) (2026)

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

- [Escape the Language Prior: Mitigating Late-Stage Modality Collapse in Audio Reasoning via Modality-Aware Policy Optimization](https://mlsanity.com/paper/2605.27741/) (2026)

## Abstract (arXiv)

Large Audio-Language Models (LALMs) have made significant progress in audio understanding, yet they primarily operate as perception-and-answer systems without explicit reasoning processes. Existing methods for enhancing audio reasoning rely either on supervised chain-of-thought (CoT) fine-tuning, which is limited by training data quality, or on reinforcement learning (RL) with coarse rewards that do not directly evaluate reasoning quality. As a result, the generated reasoning chains often appear well-structured yet lack specific acoustic grounding. We propose Audio-DeepThinker, a framework built on two core ideas. First, we introduce a hybrid reasoning similarity reward that directly supervises the quality of generated reasoning chains by combining an LLM evaluator assessing logical path alignment, key step coverage, and analytical depth with an embedding similarity component enforcing semantic alignment with reference reasoning chains. Second, we propose a progressive two-stage curriculum that enables high-quality CoT reasoning to emerge through pure RL exploration, without any supervised reasoning fine-tuning, from an instruction-tuned model that possesses no prior chain-of-thought capability. Stage 1 trains on foundational audio QA with the hybrid reward to foster basic reasoning patterns, while Stage 2 shifts to acoustically challenging boundary cases with an LLM-only reward for greater reasoning diversity. Audio-DeepThinker achieves state-of-the-art results on MMAR (74.0%), MMAU-test-mini (78.5%), and MMSU (77.26%), winning 1st Place in the Interspeech 2026 Audio Reasoning Challenge (Single Model Track). Interpretability analyses further reveal that RL training primarily reshapes upper-layer MoE gating mechanisms and that reasoning tokens crystallize progressively in the upper transformer layers, offering mechanistic insights into how audio reasoning emerges through exploration.