# Reinforcement Learning Outperforms Supervised Fine-Tuning: A Case Study on Audio Question Answering

arXiv: https://arxiv.org/abs/2503.11197 · HTML: https://mlsanity.com/paper/2503.11197/

## Key facts

- Cited by 65 papers in the map (data through 14 Oct 2026).
- Placed in Audio-language model reasoning, one of 17 directions in LLM reasoning.
- Builds on 17 of its 21 references that are in the map.
- First posted on arXiv 14 Mar 2025.

## Placed in

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

## Builds on (most cited first, 17 of 17)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [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)
- [LLaVA-CoT: Let Vision Language Models Reason Step-by-Step](https://mlsanity.com/paper/2411.10440/) (2024)
- [MMAU: A Massive Multi-Task Audio Understanding and Reasoning Benchmark](https://mlsanity.com/paper/2410.19168/) (2024)
- [R1-Zero's "Aha Moment" in Visual Reasoning on a 2B Non-SFT Model](https://mlsanity.com/paper/2503.05132/) (2025)
- [Audio Flamingo 2: An Audio-Language Model with Long-Audio Understanding and Expert Reasoning Abilities](https://mlsanity.com/paper/2503.03983/) (2025)
- [Audio-Reasoner: Improving Reasoning Capability in Large Audio Language Models](https://mlsanity.com/paper/2503.02318/) (2025)
- [Audio-CoT: Exploring Chain-of-Thought Reasoning in Large Audio Language Model](https://mlsanity.com/paper/2501.07246/) (2025)
- [All Roads Lead to Likelihood: The Value of Reinforcement Learning in Fine-Tuning](https://mlsanity.com/paper/2503.01067/) (2025)
- [SLAM-AAC: Enhancing Audio Captioning with Paraphrasing Augmentation and CLAP-Refine through LLMs](https://mlsanity.com/paper/2410.09503/) (2024)
- [Enhancing Automated Audio Captioning via Large Language Models with Optimized Audio Encoding](https://mlsanity.com/paper/2406.13275/) (2024)
- [EnCLAP++: Analyzing the EnCLAP Framework for Optimizing Automated Audio Captioning Performance](https://mlsanity.com/paper/2409.01201/) (2024)
- [Expanding on EnCLAP with Auxiliary Retrieval Model for Automated Audio Captioning](https://mlsanity.com/paper/2409.01160/) (2024)

## Built on by (most cited first, 30 of 65)

- [The Landscape of Agentic Reinforcement Learning for LLMs: A Survey](https://mlsanity.com/paper/2509.02547/) (2025)
- [MMAR: A Challenging Benchmark for Deep Reasoning in Speech, Audio, Music, and Their Mix](https://mlsanity.com/paper/2505.13032/) (2025)
- [Perception, Reason, Think, and Plan: A Survey on Large Multimodal Reasoning Models](https://mlsanity.com/paper/2505.04921/) (2025)
- [LongVT: Incentivizing "Thinking with Long Videos" via Native Tool Calling](https://mlsanity.com/paper/2511.20785/) (2025)
- [Reinforced MLLM: A Survey on RL-Based Reasoning in Multimodal Large Language Models](https://mlsanity.com/paper/2504.21277/) (2025)
- [Step-Audio-R1 Technical Report](https://mlsanity.com/paper/2511.15848/) (2025)
- [SARI: Structured Audio Reasoning via Curriculum-Guided Reinforcement Learning](https://mlsanity.com/paper/2504.15900/) (2025)
- [100 Days After DeepSeek-R1: A Survey on Replication Studies and More Directions for Reasoning Language Models](https://mlsanity.com/paper/2505.00551/) (2025)
- [STITCH: Simultaneous Thinking and Talking with Chunked Reasoning for Spoken Language Models](https://mlsanity.com/paper/2507.15375/) (2025)
- [Mini-Omni-Reasoner: Token-Level Thinking-in-Speaking in Large Speech Models](https://mlsanity.com/paper/2508.15827/) (2025)
- [Audio-Thinker: Guiding Audio Language Model When and How to Think via Reinforcement Learning](https://mlsanity.com/paper/2508.08039/) (2025)
- [Audio Flamingo Next: Next-Generation Open Audio-Language Models for Speech, Sound, and Music](https://mlsanity.com/paper/2604.10905/) (2026)
- [DRA-GRPO: Your GRPO Needs to Know Diverse Reasoning Paths for Mathematical Reasoning](https://mlsanity.com/paper/2505.09655/) (2025)
- [WavReward: Spoken Dialogue Models With Generalist Reward Evaluators](https://mlsanity.com/paper/2505.09558/) (2025)
- [Reinforcement Fine-Tuning Powers Reasoning Capability of Multimodal Large Language Models](https://mlsanity.com/paper/2505.18536/) (2025)
- [Incentivizing Consistent, Effective and Scalable Reasoning Capability in Audio LLMs via Reasoning Process Rewards](https://mlsanity.com/paper/2510.20867/) (2025)
- [Mind-Paced Speaking: A Dual-Brain Approach to Real-Time Reasoning in Spoken Language Models](https://mlsanity.com/paper/2510.09592/) (2025)
- [SageLM: A Multi-aspect and Explainable Large Language Model for Speech Judgement](https://mlsanity.com/paper/2508.20916/) (2025)
- [ThinkOmni: Lifting Textual Reasoning to Omni-modal Scenarios via Guidance Decoding](https://mlsanity.com/paper/2602.23306/) (2026)
- [AuTAgent: A Reinforcement Learning Framework for Tool-Augmented Audio Reasoning](https://mlsanity.com/paper/2602.13685/) (2026)
- [The Interspeech 2026 Audio Reasoning Challenge: Evaluating Reasoning Process Quality for Audio Reasoning Models and Agents](https://mlsanity.com/paper/2602.14224/) (2026)
- [Omni-AutoThink: Adaptive Multimodal Reasoning via Reinforcement Learning](https://mlsanity.com/paper/2512.03783/) (2025)
- [Advancing Speech Understanding in Speech-Aware Language Models with GRPO](https://mlsanity.com/paper/2509.16990/) (2025)
- [Are Audio-Language Models Listening? Audio-Specialist Heads for Adaptive Audio Steering](https://mlsanity.com/paper/2603.06854/) (2026)
- [Explore the Reinforcement Learning for the LLM based ASR and TTS system](https://mlsanity.com/paper/2509.18569/) (2025)
- [Audio-Maestro: Enhancing Large Audio-Language Models with Tool-Augmented Reasoning](https://mlsanity.com/paper/2510.11454/) (2025)
- [AV-Master: Dual-Path Comprehensive Perception Makes Better Audio-Visual Question Answering](https://mlsanity.com/paper/2510.18346/) (2025)
- [AudioRouter: Data Efficient Audio Understanding via RL based Dual Reasoning](https://mlsanity.com/paper/2602.10439/) (2026)
- [Echo: Towards Advanced Audio Comprehension via Audio-Interleaved Reasoning](https://mlsanity.com/paper/2602.11909/) (2026)
- [Audio-Zero: Label-Free Self-Evolution for Fine-Grained Audio Reasoning](https://mlsanity.com/paper/2607.20166/) (2026)

## Abstract (arXiv)

Recently, reinforcement learning (RL) has been shown to greatly enhance the reasoning capabilities of large language models (LLMs), and RL-based approaches have been progressively applied to visual multimodal tasks. However, the audio modality has largely been overlooked in these developments. Thus, we conduct a series of RL explorations in audio understanding and reasoning, specifically focusing on the audio question answering (AQA) task. We leverage the group relative policy optimization (GRPO) algorithm to Qwen2-Audio-7B-Instruct, and our experiments demonstrated state-of-the-art performance on the MMAU Test-mini benchmark, achieving an accuracy rate of 64.5%. The main findings in this technical report are as follows: 1) The GRPO algorithm can be effectively applied to large audio language models (LALMs), even when the model has only 8.2B parameters; 2) With only 38k post-training samples, RL significantly outperforms supervised fine-tuning (SFT), indicating that RL-based approaches can be effective without large datasets; 3) The explicit reasoning process has not shown significant benefits for AQA tasks, and how to efficiently utilize deep thinking remains an open question for further research; 4) LALMs still lag far behind humans auditory-language reasoning, suggesting that the RL-based approaches warrant further exploration. Our project is available at https://github.com/xiaomi-research/r1-aqa and https://huggingface.co/mispeech/r1-aqa.