# CAAD: Contrastive Audio-Aware Distillation for Efficient Speech Language Models

arXiv: https://arxiv.org/abs/2606.23052 · HTML: https://mlsanity.com/paper/2606.23052/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Speech and audio distillation, one of 12 directions in Knowledge distillation.
- Builds on 32 of its 34 references that are in the map.
- First posted on arXiv 22 Jun 2026.

## Placed in

- Knowledge distillation / [Speech and audio distillation](https://mlsanity.com/distill/speech-and-audio-distillation/)

## Builds on (most cited first, 30 of 32)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [On-Policy Distillation of Language Models: Learning from Self-Generated Mistakes](https://mlsanity.com/paper/2306.13649/) (2023)
- [Qwen2.5-Omni Technical Report](https://mlsanity.com/paper/2503.20215/) (2025)
- [VoxCeleb: a large-scale speaker identification dataset](https://mlsanity.com/paper/1706.08612/) (2017)
- [MELD: A Multimodal Multi-Party Dataset for Emotion Recognition in Conversations](https://mlsanity.com/paper/1810.02508/) (2018)
- [SpeechGPT: Empowering Large Language Models with Intrinsic Cross-Modal Conversational Abilities](https://mlsanity.com/paper/2305.11000/) (2023)
- [SALMONN: Towards Generic Hearing Abilities for Large Language Models](https://mlsanity.com/paper/2310.13289/) (2023)
- [Self-Distilled Reasoner: On-Policy Self-Distillation for Large Language Models](https://mlsanity.com/paper/2601.18734/) (2026)
- [Kimi-Audio Technical Report](https://mlsanity.com/paper/2504.18425/) (2025)
- [emotion2vec: Self-Supervised Pre-Training for Speech Emotion Representation](https://mlsanity.com/paper/2312.15185/) (2023)
- [On The Landscape of Spoken Language Models: A Comprehensive Survey](https://mlsanity.com/paper/2504.08528/) (2025)
- [PromptTTS: Controllable Text-to-Speech with Text Descriptions](https://mlsanity.com/paper/2211.12171/) (2022)
- [Dynamic-SUPERB: Towards A Dynamic, Collaborative, and Comprehensive Instruction-Tuning Benchmark for Speech](https://mlsanity.com/paper/2309.09510/) (2023)
- [Recent Advances in Speech Language Models: A Survey](https://mlsanity.com/paper/2410.03751/) (2024)
- [DailyTalk: Spoken Dialogue Dataset for Conversational Text-to-Speech](https://mlsanity.com/paper/2207.01063/) (2022)
- [The Curse of Multi-Modalities: Evaluating Hallucinations of Large Multimodal Models across Language, Visual, and Audio](https://mlsanity.com/paper/2410.12787/) (2024)
- [DeSTA2.5-Audio: Toward General-Purpose Large Audio Language Model with Self-Generated Cross-Modal Alignment](https://mlsanity.com/paper/2507.02768/) (2025)
- [DeSTA2: Developing Instruction-Following Speech Language Model Without Speech Instruction-Tuning Data](https://mlsanity.com/paper/2409.20007/) (2024)
- [MLLMs are Deeply Affected by Modality Bias](https://mlsanity.com/paper/2505.18657/) (2025)
- [DeSTA: Enhancing Speech Language Models through Descriptive Speech-Text Alignment](https://mlsanity.com/paper/2406.18871/) (2024)
- [Understanding Sounds, Missing the Questions: The Challenge of Object Hallucination in Large Audio-Language Models](https://mlsanity.com/paper/2406.08402/) (2024)
- [Reducing Object Hallucination in Large Audio-Language Models via Audio-Aware Decoding](https://mlsanity.com/paper/2506.07233/) (2025)
- [Brouhaha: multi-task training for voice activity detection, speech-to-noise ratio, and C50 room acoustics estimation](https://mlsanity.com/paper/2210.13248/) (2022)
- [Building a Taiwanese Mandarin Spoken Language Model: A First Attempt](https://mlsanity.com/paper/2411.07111/) (2024)
- [AccentDB: A Database of Non-Native English Accents to Assist Neural Speech Recognition](https://mlsanity.com/paper/2005.07973/) (2020)
- [When Audio and Text Disagree: Revealing Text Bias in Large Audio-Language Models](https://mlsanity.com/paper/2508.15407/) (2025)
- [CoLLD: Contrastive Layer-to-layer Distillation for Compressing Multilingual Pre-trained Speech Encoders](https://mlsanity.com/paper/2309.07707/) (2023)
- [CKD: Contrastive Knowledge Distillation from A Sample-wise Perspective](https://mlsanity.com/paper/2404.14109/) (2024)
- [Text Takes Over: A Study of Modality Bias in Multimodal Intent Detection](https://mlsanity.com/paper/2508.16122/) (2025)
- [How Contrastive Decoding Enhances Large Audio Language Models](https://mlsanity.com/paper/2603.09232/) (2026)

## Built on by (most cited first, 2 of 2)

- [Tri-PvP: Exposing Modality Bias in Omni-Modal Large Language Models through Perceptual-Propositional Evidence Conflicts](https://mlsanity.com/paper/2609.06011/) (2026)
- [Reward-Tilted On-Policy Distillation for Acoustic Grounding in Audio-Language Models](https://mlsanity.com/paper/2609.28778/) (2026)

## Abstract (arXiv)

Speech Language Models achieve reasoning capabilities, but are often hindered by massive parameter counts and a tendency to prioritize linguistic priors over acoustic features. While contrastive decoding enhances grounding by contrasting audio-aware and text-only logits, it increases inference latency. We propose Contrastive Audio-Aware Distillation (CAAD), a framework that internalizes the teacher's contrastive reasoning into the student model's weights. To overcome the high computational training overhead in the dual-path token-by-token contrastive distillation process, we introduce a synchronized teacher-forcing strategy. Anchored by unified Pseudo-Ground Truths, this mechanism enables simultaneous full-sequence generation of the teacher's contrastive distributions, allowing student to distill the audio-aware signal efficiently. Overall, CAAD yields a ~8% relative gain over standard knowledge distillation on Dynamic-SUPERB and successfully reduces linguistic bias in MCR-BENCH.