arXiv Machine Learning

SkillFormer: Skill-Decomposed Adaptation for Audio Language Models

SkillFormer is a method for audio language models that decomposes audio understanding into skill‑specific low‑rank adapters and uses a learned router to activate the appropriate adapters at inference time. The router selects which adapters to engage based on the question, allowing different parameters to be used for tasks such as pitch comparison versus genre classification. An alternating training schedule updates each adapter on its own skill cluster before jointly calibrating the router, reducing gradient conflicts and adding fewer than 4% of the base model’s parameters. The approach improves average accuracy by 2.5 to 4.1 points across three distinct models on MMSU, MMAU‑Pro, and MMAR, achieving balanced gains across perception, reasoning, and semantic subcategories.

arXiv AI

Oct 1

UniAE-MoE: A Unified Audio Encoder via Mixture of Experts

UniAE-MoE is a unified audio encoder that uses a Mixture-of-Experts architecture to model cross‑domain audio representations. It integrates encoder components from Qwen2‑Audio and Audio‑Flamingo 3, enhances them with SwiGLU and shared experts, and applies a two‑stage instruction‑tuning strategy along with task‑specific data scaling. The model achieves state‑of‑the‑art results on the XARES‑LLM benchmark (0.802) and tops the Interspeech 2026 Audio Encoder Capability Challenge, demonstrating strong generalization across speech, music, and general audio tasks.

By Shengbo Cai, Zhisheng Zhang, Zichao Nie, Jing Peng, Jingran Xie, Zhiyong Wu

llms benchmarks

arXiv Machine Learning

1d ago

AdaLoop: Adaptive-Depth Latent Reasoning for Audio Language Models

AdaLoop is a lightweight recurrent module that adaptively determines how many latent refinement steps are needed for audio–question pairs, allowing deeper reasoning only when necessary. It shares a transformer block that iterates over the audio representation guided by the question, with a learned halting mechanism that exits the loop once the representation is ready. Adding fewer than 3 % of the base model’s parameters, AdaLoop improves average accuracy by 2.9 to 3.8 points across three distinct models, especially on perception-heavy subtasks.

By Lee Seung-woo, Bowen Qi

llms

arXiv Computation and Language

Sep 28

Symbiotic Architecture for Post-Hoc Audio Extension of Frozen Language Models

The paper introduces a symbiotic architecture that equips large language models with audio‑understanding abilities without fine‑tuning their weights. It uses an injector module to write audio‑conditioned vectors into the LLM’s key‑value cache, allowing the model to act as an audio language model while keeping the backbone unchanged. The approach improves scalability—since injection cost depends on the injector width—and preserves the LLM’s original text performance, outperforming conventional frozen‑LLM methods and approaching fine‑tuned ALM results on audio tasks.

By Yotaro Kubo, Qi Sun, Yujin Tang

llms nlp fine-tuning multimodal

arXiv AI

Jul 14

MUGEN: Evaluating and Improving Multi-audio Understanding of Large Audio-Language Models

arXiv:2603. 09714v2 Announce Type: replace-cross Abstract: While multi-audio understanding is critical for large audio-language models (LALMs), it remains underexplored.

By Chih-Kai Yang, Yun-Shao Tsai, Yu-Kai Guo, Ping-Le Tsai, Yen-Ting Piao, Hung-Wei Chen, Ting-Lin Hsiao, Yun-Man Hsu, Ke-Han Lu, Hung-yi Lee

llms benchmarks

arXiv AI

Jun 15

AudioDER: A Deduplication-Enhanced Reasoning Dataset for Post-Training Large Audio-Language Models

arXiv:2606. 14591v1 Announce Type: cross Abstract: Large Audio-Language Models (LALMs) have shown strong performance on a wide range of audio understanding tasks, yet they still struggle with complex audio reasoning.

By Hui Geng, Yi Su, Han Yin, Tianjiao Wan, Qisheng Xu, Jiaxin Chen, Zijian Gao, Hengzhu Liu, Xie Chen, Kele Xu

llms benchmarks

arXiv Machine Learning

Sep 22

ParA-LLM: A Unified Approach to Paralinguistic and Acoustic Speech Understanding

arXiv:2609.22771v1 Announce Type: cross Abstract: Recent advances in Audio LLMs have achieved human-level speech recognition, yet existing systems struggle to capture paralinguistic aspects such as s...

By Nishit Anand, Jiaqi Su, Ke Chen, Yunyun Wang, Dinesh Manocha, Ramani Duraiswami, Rithesh Kumar, Zeyu Jin

llms multimodal benchmarks

arXiv AI

Jun 4

Audio Interaction Model

arXiv:2606. 05121v1 Announce Type: cross Abstract: Audio is an inherently interactive modality, yet today's Large Audio Language Models (LALMs) are offline, and streaming audio models each handle only a single task such as streaming ASR or voice chatting.

By Zhifei Xie, Zihang Liu, Ze An, Xiaobin Hu, Yue Liao, Ziyang Ma, Dongchao Yang, Mingbao Lin, Deheng Ye, Shuicheng Yan, Chunyan Miao

llms multimodal benchmarks

arXiv AI

Jun 2

MOSS-Audio Technical Report

arXiv:2606. 01802v1 Announce Type: cross Abstract: MOSS-Audio is a unified audio-language model for speech, environmental sound, and music understanding, supporting audio captioning, time-aware question answering, timestamped transcription, and audio-grounded reasoning.

By Chen Yang, Chufan Yu, Hanfu Chen, Jie Zhu, Jingqi Chen, Ke Chen, Wenxuan Wang, Yang Wang, Yaozhou Jiang, Yi Jiang, Zhengyuan Lin, Ziqi Chen, Zhaoye Fei, Chenghao Liu, Jun Zhan, Kang Yu, Kexin Huang, Mingshu Chen, Qinyuan Cheng, Ruixiao Li, Shimin Li, Songlin Wang, Yang Gao, Yiyang Zhang, Xipeng Qiu

llms agents nlp fine-tuning multimodal

arXiv AI

Jul 14

Encoder-Side Neuron Identification and Amplification for Acoustic Perception in Large Audio-Language Models

arXiv:2607. 11801v1 Announce Type: cross Abstract: Large audio-language models (LALMs) often underperform on fine-grained, non-semantic attributes of speech, such as a speaker's emotion, despite strong performance on speech content.

By Yu-Han Huang, Chih-Kai Yang, Ke-Han Lu, An-Yu Cheng, Hung-yi Lee

llms fine-tuning

arXiv AI

Jul 1

ALM2Vec: Learning Audio Embeddings for Universal Audio Retrieval with Large Audio-Language Models

arXiv:2606. 30682v1 Announce Type: cross Abstract: Recent advances in language--audio retrieval have been largely driven by contrastive dual-encoder architectures that align audio and text in a shared embedding space.

By Fengjie Lu, Chenang Jiang, Jiarui Hai, Helin Wang, Aaron Yee

llms rag nlp multimodal benchmarks

arXiv AI

Jun 18

Continuous Audio Thinking for Large Audio Language Models

arXiv:2606. 18273v1 Announce Type: cross Abstract: Large audio language models (LALMs) have shown impressive capabilities on diverse audio understanding tasks, ranging from speech transcription to music analysis.

By Gyojin Han, Dong-Jae Lee, Changho Choi, Jongsuk Kim, Junmo Kim

llms efficiency benchmarks

arXiv Computation and Language

Sep 1

A Simple Method to Enhance Pre-trained Language Models with Speech Tokens for Classification

arXiv:2512.07571v3 Announce Type: replace Abstract: This paper presents a simple method that allows to easily enhance textual pre-trained large language models with speech information, when fine-tune...

By Nicolas Calbucura, Jose Guillen, Valentin Barriere

llms rag nlp fine-tuning multimodal

Related stories