Continuous Audio Thinking for Large Audio Language Models

• URL: http://arxiv.org/abs/2606.18273v1
• Date: Fri, 05 Jun 2026 11:38:30 GMT
• Title: Continuous Audio Thinking for Large Audio Language Models
• Abstract summary: We introduce Continuous Audio Thinking (CoAT) to equip audio language models with a continuous latent workspace for organizing acoustic information prior to response generation.<n>Within the thinking space, the model can utilize the rich acoustic information provided by expert distillation when generating its response.<n>CoAT does not require additional autoregressive decoding cost over the baseline.
• Score: 16.335310406868217
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large audio language models (LALMs) have shown impressive capabilities on diverse audio understanding tasks, ranging from speech transcription to music analysis. However, because LALMs are typically trained to produce text-aligned responses, their hidden states are progressively shaped for text generation rather than for preserving acoustic information. As a result, the diverse acoustic content that audio carries, such as phonetic detail, prosody, sound events, affect, and pitch, is lost along the way and difficult to leverage in the response. We introduce Continuous Audio Thinking (CoAT), a framework that equips audio language models with a continuous latent workspace for organizing acoustic information prior to response generation, grounded by distillation from audio experts. Within the thinking space, the model can utilize the rich acoustic information provided by expert distillation when generating its response. Furthermore, the proposed continuous thinking block can be processed in a single prefill, so CoAT does not require additional autoregressive decoding cost over the baseline. Across three LALMs, Qwen2-Audio, Qwen2.5-Omni-7B, and Audio Flamingo~3, performance gains on a broad benchmark suite spanning audio reasoning, audio understanding, music classification, speech emotion, and speech transcription demonstrate the effectiveness of CoAT. Further analysis confirms that the auxiliary supervision propagates from the thinking positions to the model's textual responses.

Related papers

• FireRedAudio: A General-Purpose Audio Language Model with Decoupled Continuous Representations for Understanding and Generation [46.93738546805678]
  A unified audio model must recognize and understand linguistic, paralinguistic, and environmental information while supporting speech synthesis and editing.<n>We introduce FireRedAudio, a general-purpose audio language model with a shared 9B- parameter LLM.<n>FireRedAudio achieves competitive or leading performance in audio understanding and multilingual ASR, strong content accuracy and speaker timestamp preservation in zero-shot TTS, and substantial improvements over Ming-UniAudio-Edit.
  arXiv  Detail & Related papers  (2026-08-25T07:37:15Z)
• AudioDER: A Deduplication-Enhanced Reasoning Dataset for Post-Training Large Audio-Language Models [42.62457130960257]
  Large Audio-Language Models (LALMs) have shown strong performance on a wide range of audio understanding tasks, yet they still struggle with complex audio reasoning.<n>Existing audio-language datasets often contain substantial redundancy, where many samples are highly similar in acoustic content.<n>We propose a redundancy-aware data construction pipeline for building reasoning-oriented supervision for LALMs.
  arXiv  Detail & Related papers  (2026-06-12T16:09:49Z)
• AudioMosaic: Contrastive Masked Audio Representation Learning [53.52371029884106]
  We introduce textbfAudioMosaic, a contrastive learning-based audio encoder for general audio understanding.<n>AudioMosaic constructs positive pairs by applying structured time-frequency masking to spectrogram patches.<n>Experiments show that AudioMosaic achieves state-of-the-art performance on several standard audio benchmarks.
  arXiv  Detail & Related papers  (2026-05-14T00:56:51Z)
• Echo: Towards Advanced Audio Comprehension via Audio-Interleaved Reasoning [39.264735719707154]
  Current efforts replicate text-based reasoning by contextualizing audio content through a one-time encoding.<n>We propose audio-interleaved reasoning to break through this bottleneck.<n>We present Echo, a LALM capable of dynamically re-listening to audio in demand during reasoning.
  arXiv  Detail & Related papers  (2026-02-12T13:06:34Z)
• Beyond Transcripts: A Renewed Perspective on Audio Chaptering [66.61445564139052]
  We show that a novel audio-only architecture (AudioSeg) outperforms text-based approaches for segmenting long-form audio into coherent sections.<n>Our experiments on YTSeg reveal that AudioSeg substantially outperforms text-based approaches, pauses provide the largest acoustic gains, and MLLMs remain limited by context length and weak instruction following.
  arXiv  Detail & Related papers  (2026-02-09T18:28:10Z)
• ThinkSound: Chain-of-Thought Reasoning in Multimodal Large Language Models for Audio Generation and Editing [47.14083940177122]
  ThinkSound is a novel framework that enables stepwise, interactive audio generation and editing for videos.<n>Our approach decomposes the process into three complementary stages: semantically coherent, interactive object-centric refinement, and targeted editing.<n> Experiments demonstrate that ThinkSound achieves state-of-the-art performance in video-to-audio generation across both audio metrics and CoT metrics.
  arXiv  Detail & Related papers  (2025-06-26T16:32:06Z)
• From Alignment to Advancement: Bootstrapping Audio-Language Alignment with Synthetic Data [55.2480439325792]
  Audio-aware large language models (ALLMs) have recently made great strides in understanding and processing audio inputs.<n>These models are typically adapted from text-based large language models (LLMs) through additional training on audio-related tasks.<n>We propose a data generation framework that produces contrastive-like training data, designed to enhance ALLMs' ability to differentiate between present and absent sounds.
  arXiv  Detail & Related papers  (2025-05-26T16:08:41Z)
• ADIFF: Explaining audio difference using natural language [31.963783032080993]
  This paper comprehensively studies the task of explaining audio differences and then propose benchmark, baselines for the task.<n>We present two new datasets for audio difference explanation derived from the AudioCaps and Clotho audio captioning datasets.<n>We propose ADIFF, which introduces a cross-projection module, position captioning, and a three-step training process to enhance the model's ability to produce detailed explanations.
  arXiv  Detail & Related papers  (2025-02-06T20:00:43Z)
• Enhancing Temporal Understanding in Audio Question Answering for Large Audio Language Models [0.9285295512807729]
  The Audio Question Answering (AQA) task includes audio event classification, audio captioning, and open-ended reasoning.<n>LALMs excel in general audio understanding, but are limited in temporal reasoning.<n>This paper addresses these challenges and limitations in audio temporal reasoning.
  arXiv  Detail & Related papers  (2024-09-10T05:26:53Z)
• Tango 2: Aligning Diffusion-based Text-to-Audio Generations through Direct Preference Optimization [70.13218512896032]
  Generation of audio from text prompts is an important aspect of such processes in the music and film industry. Our hypothesis is focusing on how these aspects of audio generation could improve audio generation performance in the presence of limited data. We synthetically create a preference dataset where each prompt has a winner audio output and some loser audio outputs for the diffusion model to learn from.
  arXiv  Detail & Related papers  (2024-04-15T17:31:22Z)
• Separate Anything You Describe [53.30484933564858]
  Language-queried audio source separation (LASS) is a new paradigm for computational auditory scene analysis (CASA)<n>AudioSep is a foundation model for open-domain audio source separation with natural language queries.
  arXiv  Detail & Related papers  (2023-08-09T16:09:44Z)
• AudioLM: a Language Modeling Approach to Audio Generation [59.19364975706805]
  We introduce AudioLM, a framework for high-quality audio generation with long-term consistency. We show how existing audio tokenizers provide different trade-offs between reconstruction quality and long-term structure. We demonstrate how our approach extends beyond speech by generating coherent piano music continuations.
  arXiv  Detail & Related papers  (2022-09-07T13:40:08Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.