Escape the Language Prior: Mitigating Late-Stage Modality Collapse in Audio Reasoning via Modality-Aware Policy Optimization

• URL: http://arxiv.org/abs/2605.27741v1
• Date: Tue, 26 May 2026 22:34:03 GMT
• Title: Escape the Language Prior: Mitigating Late-Stage Modality Collapse in Audio Reasoning via Modality-Aware Policy Optimization
• Abstract summary: We introduce Modality-Aware Policy Optimization (MAPO), a novel dual-branch reinforcement learning framework.<n>First, MAPO dynamically concentrates the policy gradient on modality-critical tokens.<n>Second, it integrates an auxiliary attention loss branch that applies a targeted, temporally scaled penalty to the model's internal attention distributions.
• Score: 40.86280811828235
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Audio and omni-modal large language models exhibit impressive cross-modal reasoning capabilities. However, applying standard reinforcement learning post-training algorithms to these models exposes a critical structural vulnerability: methods like GRPO apply uniform policy gradients across all tokens, ignoring their unequal dependence on the non-text source modality. This exacerbates late-stage modality collapse during extended chain-of-thought generation, where models progressively abandon the primary source signal in favor of compressed textual priors, leading to confident but ungrounded hallucinations. To address this, we introduce Modality-Aware Policy Optimization (MAPO), a novel dual-branch reinforcement learning framework. First, MAPO dynamically concentrates the policy gradient on modality-critical tokens using a modality relevance mask, which is derived from the cross-modal differential entropy between an audio-ablated reference and the multimodal policy. Second, it integrates an auxiliary attention loss branch that applies a targeted, temporally scaled penalty to the model's internal attention distributions. This ensures the model actively sustains cross-modal grounding deep into the reasoning trace. Evaluations on complex audio reasoning benchmarks demonstrate that MAPO substantially improves long-horizon reasoning fidelity and multimodal instruction following, achieving highly competitive performance and setting new state-of-the-art results on several key benchmarks among open-weight models. By relying strictly on native statistical signals rather than domain-specific inductive biases, MAPO offers a promising foundation for mitigating epistemic collapse across diverse multimodal systems.

Related papers

• Which Modality Decides? Counterfactual Modality Attribution for Multimodal LLMs [4.192318018134786]
  Multimodal large language models (MLLMs) increasingly support high-stakes decision making by combining complementary information from images and text.<n>Existing explainability methods cannot answer a fundamental question: which modality drives a prediction?<n>We propose Counterfactual Modality Attribution (CMA), the first framework for quantifying modality-level contributions in MLLMs.
  arXiv  Detail & Related papers  (2026-07-29T13:33:11Z)
• N-GRPO: Embedding-Level Neighbor Mixing for Enhanced Policy Optimization [55.14402862283128]
  We introduce N-GRPO, a novel exploration strategy integrated into the Group Relative Policy Optimization framework.<n>Rather than relying on token-level sampling or native embedding-level noise, our approach leverages Semantic Neighbor Mixing.<n>N-GRPO achieves consistent improvements over strong baselines on math reasoning benchmarks and also exhibits robust generalization capabilities on out-of-distribution tasks.
  arXiv  Detail & Related papers  (2026-06-09T12:21:27Z)
• Enhancing Multimodal In-Context Learning via Inductive-Deductive Reasoning [65.15766304205657]
  In-context learning (ICL) allows large models to adapt to tasks using a few examples, yet its extension to vision-language models (VLMs) remains fragile.<n>Our analysis reveals that the fundamental limitation lies in an inductive gap, models often produce correct answers from flawed reasoning.<n>We introduce a framework that restructures multimodal ICL as a principled inductive-deductive process.
  arXiv  Detail & Related papers  (2026-05-04T09:18:19Z)
• Evaluation Before Generation: A Paradigm for Robust Multimodal Sentiment Analysis with Missing Modalities [21.767502810187477]
  The missing modality problem poses a fundamental challenge in multimodal sentiment analysis.<n>Existing approaches primarily improve robustness through prompt learning and pre trained models.<n>A Prompt based Missing Modality Adaptation framework is proposed to address these issues.
  arXiv  Detail & Related papers  (2026-04-07T07:59:06Z)
• Not All Tokens See Equally: Perception-Grounded Policy Optimization for Large Vision-Language Models [38.47027398567909]
  Perception-Grounded Policy Optimization (PGPO) is a novel fine-grained credit assignment framework that dynamically reshapes advantages at the token level.<n>We show that PGPO actively amplifies learning signals for visually-dependent tokens while suppressing gradient noise from linguistic priors.<n>Both theoretical and empirical analyses confirm that PGPO effectively reduces gradient variance, prevents training collapse, and acts as a potent regularizer for robust, perception-grounded multimodal reasoning.
  arXiv  Detail & Related papers  (2026-04-02T09:53:20Z)
• HIPO: Instruction Hierarchy via Constrained Reinforcement Learning [57.40686733111483]
  textscHIPO is a novel alignment framework that formulates HIF as a Constrained Markov Decision Process.<n>textscHIPO elevates system prompts from mere input context to strict algorithmic boundaries.
  arXiv  Detail & Related papers  (2026-03-17T06:12:41Z)
• MoD-DPO: Towards Mitigating Cross-modal Hallucinations in Omni LLMs using Modality Decoupled Preference Optimization [4.088161686930475]
  We propose Modality-Decoupled Direct Preference Optimization (MoD-DPO) for improving modality grounding in omni LLMs.<n>MoD-DPO introduces modality-aware regularization terms that explicitly enforce invariance to corruptions in irrelevant modalities and sensitivity to perturbations in relevant modalities.<n>Experiments demonstrate that MoD-DPO consistently improves perception accuracy and hallucination resistance, outperforming previous preference optimization baselines.
  arXiv  Detail & Related papers  (2026-03-03T17:50:24Z)
• Multi-Path Collaborative Reasoning via Reinforcement Learning [54.8518809800168]
  Chain-of-Thought (CoT) reasoning has significantly advanced the problem-solving capabilities of Large Language Models (LLMs)<n>Recent methods attempt to address this by generating soft abstract tokens to enable reasoning in a continuous semantic space.<n>We propose Multi-Path Perception Policy Optimization (M3PO), a novel reinforcement learning framework that explicitly injects collective insights into the reasoning process.
  arXiv  Detail & Related papers  (2025-12-01T10:05:46Z)
• Improving Multimodal Sentiment Analysis via Modality Optimization and Dynamic Primary Modality Selection [54.10252086842123]
  Multimodal Sentiment Analysis (MSA) aims to predict sentiment from language, acoustic, and visual data in videos.<n>This paper proposes a modality optimization and dynamic primary modality selection framework (MODS)<n>Experiments on four benchmark datasets demonstrate that MODS outperforms state-of-the-art methods.
  arXiv  Detail & Related papers  (2025-11-09T11:13:32Z)

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.