Enregistré dans:
Détails bibliographiques
Auteurs principaux: Wang, Juncheng, Hu, Zhe, Xu, Chao, Ren, Siyue, Feng, Yuxiang, Liu, Yang, Sun, Baigui, Wang, Shujun
Format: Preprint
Publié: 2026
Sujets:
Accès en ligne:https://arxiv.org/abs/2601.14304
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866909996422266880
author Wang, Juncheng
Hu, Zhe
Xu, Chao
Ren, Siyue
Feng, Yuxiang
Liu, Yang
Sun, Baigui
Wang, Shujun
author_facet Wang, Juncheng
Hu, Zhe
Xu, Chao
Ren, Siyue
Feng, Yuxiang
Liu, Yang
Sun, Baigui
Wang, Shujun
contents Autoregressive (AR) models excel at generating temporally coherent audio by producing tokens sequentially, yet they often falter in faithfully following complex textual prompts, especially those describing complex sound events. We uncover a surprising capability in AR audio generators: their early prefix tokens implicitly encode global semantic attributes of the final output, such as event count and sound-object category, revealing a form of implicit planning. Building on this insight, we propose Plan-Critic, a lightweight auxiliary model trained with a Generalized Advantage Estimation (GAE)-inspired objective to predict final instruction-following quality from partial generations. At inference time, Plan-Critic enables guided exploration: it evaluates candidate prefixes early, prunes low-fidelity trajectories, and reallocates computation to high-potential planning seeds. Our Plan-Critic-guided sampling achieves up to a 10-point improvement in CLAP score over the AR baseline-establishing a new state of the art in AR text-to-audio generation-while maintaining computational parity with standard best-of-N decoding. This work bridges the gap between causal generation and global semantic alignment, demonstrating that even strictly autoregressive models can plan ahead.
format Preprint
id arxiv_https___arxiv_org_abs_2601_14304
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Guided by the Plan: Enhancing Faithful Autoregressive Text-to-Audio Generation with Guided Decoding
Wang, Juncheng
Hu, Zhe
Xu, Chao
Ren, Siyue
Feng, Yuxiang
Liu, Yang
Sun, Baigui
Wang, Shujun
Computation and Language
Sound
Audio and Speech Processing
Autoregressive (AR) models excel at generating temporally coherent audio by producing tokens sequentially, yet they often falter in faithfully following complex textual prompts, especially those describing complex sound events. We uncover a surprising capability in AR audio generators: their early prefix tokens implicitly encode global semantic attributes of the final output, such as event count and sound-object category, revealing a form of implicit planning. Building on this insight, we propose Plan-Critic, a lightweight auxiliary model trained with a Generalized Advantage Estimation (GAE)-inspired objective to predict final instruction-following quality from partial generations. At inference time, Plan-Critic enables guided exploration: it evaluates candidate prefixes early, prunes low-fidelity trajectories, and reallocates computation to high-potential planning seeds. Our Plan-Critic-guided sampling achieves up to a 10-point improvement in CLAP score over the AR baseline-establishing a new state of the art in AR text-to-audio generation-while maintaining computational parity with standard best-of-N decoding. This work bridges the gap between causal generation and global semantic alignment, demonstrating that even strictly autoregressive models can plan ahead.
title Guided by the Plan: Enhancing Faithful Autoregressive Text-to-Audio Generation with Guided Decoding
topic Computation and Language
Sound
Audio and Speech Processing
url https://arxiv.org/abs/2601.14304