Model Releases
TAD: Token-Adaptive Contrastive Decoding with Confidence-Guided Gating for Hallucination Mitigation in Large Audio-Language Models
arXiv:2609.07286v1 Announce Type: cross Abstract: Large audio-language models (LALMs) can hallucinate audio objects, answering 'yes' to absent sound events, thus undermining reliability in audio quest
arXiv:2609.07286v1 Announce Type: cross Abstract: Large audio-language models (LALMs) can hallucinate audio objects, answering "yes" to absent sound events, thus undermining reliability in audio question answering. We propose Token-Adaptive Decoding (TAD), a training-free strategy for hallucination mitigation that grounds the initial yes/no decision by contrasting logits under real audio with a matched silent reference. TAD introduces a token-adaptive, confidence-guided gate that is decision-critical at the first decoding step and class-conditional on affirmative tokens, using the audio-silent margin to avoid overcorrection when evidence is weak or already sufficient. Experiments on AudioCaps-Hallucination show that, relative to Audio-Aware Decoding (AAD), a contrastive baseline with fixed contrast strength, TAD improves F1 for Qwen2 by 0.059 to 0.117 across Popular, Adversarial, and Random splits, and for Gemma by 0.025 to 0.064, while on Clotho-AQA it raises F1 from 0.810 to 0.816 on Qwen2 and remains comparable to AAD on Gemma.
Related
- MRMAD: A Multi-Round Multi-Audio Benchmark for Evaluating Acoustic Degradation Perception in Large Audio-Language Models
- Does Playing it Safe Count as Faithfulness? Reassessing LVLM Hallucination Mitigation Methods
- TPCD: Tone-Pressure Contrastive Decoding and the Label-Free Gating Bottleneck in Vision-Language Models
Source: arXiv cs.CV | 2026-09-09