{"slug": "x-aut-progressive-audio-encoder-compression-for-speech-llms-with-cross-scale", "title": "X-AuT: Progressive Audio-Encoder Compression for Speech LLMs with Cross-Scale Distillation", "summary": "Researchers introduced X-AuT, a progressive framework that selects layer combinations to compress audio-encoder depth for speech large language models, addressing the deletion and premature end-of-sequence errors caused by removing complete blocks. The method uses cross-scale distillation to reduce inference cost while limiting perturbation to the embeddings consumed by the decoder.", "body_md": "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 combination", "url": "https://wpnews.pro/news/x-aut-progressive-audio-encoder-compression-for-speech-llms-with-cross-scale", "canonical_source": "https://aiflash.com/news/117455/", "published_at": "2026-09-11 06:30:13+00:00", "updated_at": "2026-09-11 06:56:26.944124+00:00", "lang": "en", "topics": ["large-language-models", "ai-research", "natural-language-processing"], "entities": ["X-AuT"], "alternates": {"html": "https://wpnews.pro/news/x-aut-progressive-audio-encoder-compression-for-speech-llms-with-cross-scale", "markdown": "https://wpnews.pro/news/x-aut-progressive-audio-encoder-compression-for-speech-llms-with-cross-scale.md", "text": "https://wpnews.pro/news/x-aut-progressive-audio-encoder-compression-for-speech-llms-with-cross-scale.txt", "jsonld": "https://wpnews.pro/news/x-aut-progressive-audio-encoder-compression-for-speech-llms-with-cross-scale.jsonld"}}