Continuous Audio Thinking for Large Audio Language Models

arXiv:2606.1827317.9
Predicted impact top 50% in CL · last 90 daysOriginality Incremental advance
AI Analysis

For developers of audio-language models, CoAT addresses the bottleneck of acoustic information loss during text generation, offering a method to improve performance on audio understanding tasks.

Large audio language models lose acoustic information during text generation. The proposed Continuous Audio Thinking (CoAT) framework introduces a continuous latent workspace with expert distillation to preserve acoustic details, achieving performance gains across multiple LALMs on diverse audio tasks without extra decoding cost.

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.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes