{
  "id": 4588,
  "url": "https://arxiv.org/abs/2605.23954v1",
  "title": "EchoDistill:Alignment Noisy-to-Clean Self-Distillation for Robust Audio LLMs",
  "summary": "Audio Large Language Models (ALLMs) are highly vulnerable to real-world noise, which often induces severe semantic drift and hallucinations. Existing robustness methods primarily rely on waveform-level acoustic enhancement, answer-level supervision, or the internal suppression of noise representations. To address these issues, we propose echodistill, an alignment-based noisy-to-clean self-distillation framework. Echodistill leverages a frozen clean-audio teacher to provide semantic references fo",
  "authors": "Liang Lin, Chunxi Luo, Kaiwen Luo, Jie Zhang, Jin Wang, Yuanhe Zhang et al.",
  "category": "research",
  "topics": "safety-alignment",
  "orgs": null,
  "regions": null,
  "published_at": "2026-05-11T06:30:25.000Z",
  "fetched_at": "2026-07-14T16:31:08.354Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/4588",
  "original_url": "https://arxiv.org/abs/2605.23954v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}