arxiv:2609.11412
Published on Sep 10
· Submitted by
Shiyu Huang on Sep 11
· XPENG AI
Upvote
46
Authors:
,
,
,
Qize Yu ,
,
,
,
,
,
Abstract
X-AuT progressively prunes audio-encoder layers in speech large language models and restores accuracy via behavioral probes, representation alignment, cross-scale distillation, and LoRA adaptation.
Generated by thinkingmachines/Inkling-Small
Reducing audio-encoder depth lowers the inference cost of speech large language models, but removing complete blocks perturbs the embeddings consumed by the decoder and can cause deletion and premature end-of-sequence errors. We introduce X-AuT, a progressive framework that selects layer combinations through short behavioral probes and restores the pruned model through representation alignment, cross-scale distillation, scheduled student-policy supervision, and LoRA finetuning. The language-model backbone remains frozen, while attention LoRA adapters and the tied output embedding adapt during distillation. Training uses the highest-agreement tier from a transcript-consistency pipeline, followed by source reweighting during finetuning. On ten public Chinese--English benchmarks, compressing Qwen3-ASR-0.6B from 18 to 16 audio-encoder layers reduces macro-average error from 5.61% to 5.27%. The 14-layer model reaches 5.75% with 20.7% fewer audio-tower parameters. Under the matched recipe, the 1.7B teacher yields 5.55% mean error, compared with 8.45% for self-distillation, and progressive 18rightarrow14 pruning outperforms direct pruning (5.75% vs. 6.73%). These single-run results establish two practical operating points and show that the accuracy effects vary across benchmarks. Project website: https://xpeng-ai.github.io/[x-aut](https://hf-mirror.com/papers?q=x-aut)
View arXiv page View PDF Project pageGitHub 11 Add to collection
Community
Paper author Paper submitter 12 days ago
Project website: xpeng-ai.github.io/x-aut Paper: arXiv:2609.11412 Source code: XPENG-AI/X-AuT Model weights: XPENG-AI/X-AuT
11 days ago
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
-
Compressing Streaming Neural Audio Encoders via Latent-Space Distillation (2026)
-
Where Does the Sound Go? Tracing Acoustic Information Loss in Audio-Conditioned LLMs (2026)
-
Qwen-MusicAVQA-7B: A Multimodal Model for Music Audio-Visual QA (2026)
-
Stride-k Subsampling: Train-Free Audio Token Reduction for Whisper (2026)
-
ParaASR: Multi-Token Prediction for Fast and Long-Context LLM-Based Speech Recognition (2026)
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on HF Mirror checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: @librarian-bot recommend
8 days ago
This is an automated message from the ResearchStudio team.
We created an interactive ResearchStudio Reel for this paper. It includes a visual poster, a video, and a blog, all available for download in editable formats.
Open the ResearchStudio Reel →
Download all files from HF Mirror
Please give this comment a thumbs up if you find the Reel helpful!
Want to explore or create Reels for more papers? Visit the ResearchStudio demo.
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
· Sign up or log in to comment
Upvote
46
Models citing this paper 1
Datasets citing this paper 0
No dataset linking this paper
Cite arxiv.org/abs/2609.11412 in a dataset README.md to link it from this page.