# Temporal Contrastive Decoding: A Training-Free Method for Large Audio-Language Models

arXiv: https://arxiv.org/abs/2604.15383 · HTML: https://mlsanity.com/paper/2604.15383/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 33 references that are in the map.
- First posted on arXiv 16 Apr 2026.

## Builds on (most cited first, 21 of 21)

- [Plug and Play Language Models: A Simple Approach to Controlled Text Generation](https://mlsanity.com/paper/1912.02164/) (2019)
- [Mitigating Object Hallucinations in Large Vision-Language Models through Visual Contrastive Decoding](https://mlsanity.com/paper/2311.16922/) (2023)
- [Contrastive Decoding: Open-ended Text Generation as Optimization](https://mlsanity.com/paper/2210.15097/) (2022)
- [SpeechGPT: Empowering Large Language Models with Intrinsic Cross-Modal Conversational Abilities](https://mlsanity.com/paper/2305.11000/) (2023)
- [GeDi: Generative Discriminator Guided Sequence Generation](https://mlsanity.com/paper/2009.06367/) (2020)
- [MMAU: A Massive Multi-Task Audio Understanding and Reasoning Benchmark](https://mlsanity.com/paper/2410.19168/) (2024)
- [Mini-Omni: Language Models Can Hear, Talk While Thinking in Streaming](https://mlsanity.com/paper/2408.16725/) (2024)
- [Audio Flamingo: A Novel Audio Language Model with Few-Shot Learning and Dialogue Abilities](https://mlsanity.com/paper/2402.01831/) (2024)
- [GAMA: A Large Audio-Language Model with Advanced Audio Understanding and Complex Reasoning Abilities](https://mlsanity.com/paper/2406.11768/) (2024)
- [Audio Flamingo 2: An Audio-Language Model with Long-Audio Understanding and Expert Reasoning Abilities](https://mlsanity.com/paper/2503.03983/) (2025)
- [Decoding-Time Language Model Alignment with Multiple Objectives](https://mlsanity.com/paper/2406.18853/) (2024)
- [Clotho-AQA: A Crowdsourced Dataset for Audio Question Answering](https://mlsanity.com/paper/2204.09634/) (2022)
- [MiMo-Audio: Audio Language Models are Few-Shot Learners](https://mlsanity.com/paper/2512.23808/) (2025)
- [Slow-Fast Auditory Streams For Audio Recognition](https://mlsanity.com/paper/2103.03516/) (2021)
- [Self-Correcting Decoding with Generative Feedback for Mitigating Hallucinations in Large Vision-Language Models](https://mlsanity.com/paper/2502.06130/) (2025)
- [CochlScene: Acquisition of acoustic scene data using crowdsourcing](https://mlsanity.com/paper/2211.02289/) (2022)
- [AVCD: Mitigating Hallucinations in Audio-Visual Large Language Models through Contrastive Decoding](https://mlsanity.com/paper/2505.20862/) (2025)
- [Reducing Object Hallucination in Large Audio-Language Models via Audio-Aware Decoding](https://mlsanity.com/paper/2506.07233/) (2025)
- [Mitigating Hallucinations in Large Vision-Language Models (LVLMs) via Language-Contrastive Decoding (LCD)](https://mlsanity.com/paper/2408.04664/) (2024)
- [Who Can Withstand Chat-Audio Attacks? An Evaluation Benchmark for Large Audio-Language Models](https://mlsanity.com/paper/2411.14842/) (2024)
- [Delta -- Contrastive Decoding Mitigates Text Hallucinations in Large Language Models](https://mlsanity.com/paper/2502.05825/) (2025)

## Built on by (most cited first, 3 of 3)

- [Adaptive Perturbation Selection for Contrastive Audio Decoding](https://mlsanity.com/paper/2607.00247/) (2026)
- [Encoder-Side Neuron Identification and Amplification for Acoustic Perception in Large Audio-Language Models](https://mlsanity.com/paper/2607.11801/) (2026)
- [Does Location Matter? A Black-Box Causal Audit Reveals Model-Dependent Audio Grounding in LALM Hallucination](https://mlsanity.com/paper/2610.13112/) (2026)

## Abstract (arXiv)

Large audio-language models (LALMs) generalize across speech, sound, and music, but unified decoders can exhibit a temporal smoothing bias: transient acoustic cues may be underutilized in favor of temporally smooth context that is better supported by language priors, leading to less specific audio-grounded outputs. We propose Temporal Contrastive Decoding (TCD), a training-free decoding method for unified LALMs that mitigates this effect at inference time. TCD constructs a temporally blurred slow-path view by smoothing the input waveform and re-encoding it, then contrasts next-token logits from the original and slow-path views. The contrastive signal is applied as a token-level logit update restricted to a small candidate set. A self-normalized stability score sets the blur window and update scale, and a step-wise gate based on uncertainty and audio reliance activates the update only when needed. Experiments on MMAU and AIR-Bench show consistent improvements on strong unified LALMs. We further conduct ablations and an architectural applicability study to analyze the contributions of key components and how TCD behaves across large audio-language model designs.