# Noise-Aware In-Context Learning for Hallucination Mitigation in ALLMs

Source: [Noise-Aware In-Context Learning for Hallucination Mitigation in ALLMs](<https://www.sota2.com/research/paper/69dc78617d0172140ec7891c>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-14T05:00:04.234Z

Authors: Qixuan Huang, Khalid Zaman, Masashi Unoki

Published: 2026-04-10T06:35:46.000Z

## Abstract

Auditory large language models (ALLMs) have demonstrated strong general capabilities in audio understanding and reasoning tasks. However, their reliability is still undermined by hallucination issues. Existing hallucination evaluation methods are formulated as binary classification tasks, which are insufficient to characterize the more complex hallucination patterns that arise in generative tasks. Moreover, current hallucination mitigation strategies rely on fine-tuning, resulting in high computational costs. To address the above limitations, we propose a plug-and-play Noise-Aware In-Context Learning (NAICL) method. Specifically, we construct a noise prior library, retrieve noise examples relevant to the input audio, and incorporate them as contextual priors, thereby guiding the model to reduce speculative associations when acoustic evidence is insufficient and to adopt a more conservative generation strategy. In addition, we establish a hallucination benchmark for audio caption tasks including the construction of the Clotho-1K multi-event benchmark dataset, the definition of four types of auditory hallucinations, and the introduction of metrics such as hallucination type distribution to support fine-grained analysis. Experimental results show that all evaluated ALLMs exhibit same hallucination behaviors. Moreover, the proposed NAICL method reduces the overall hallucination rate from 26.53% to 16.98%.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Audio Hallucination Evaluation on Clotho-1K](<https://www.sota2.com/research/sota/audio-hallucination-evaluation-on-clotho-1k.md>): Audio Hallucination Evaluation; Clotho-1K; HR: 16.98.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2604.09021>)

[PDF](<https://arxiv.org/pdf/2604.09021.pdf>)