LLM Digest

Story

arxiv_llm_reliability · Sep 30, 2026 · paper

Source brief

AnchorPrompt: Self-Distilled Soft Prompts for Robust Audio-Language Models

arxiv.orgSep 30, 2026
original source linked

In brief

Large audio-language models (LALMs) are sensitive to input perturbations, such as noise, waveform corruption, and adversarial injections. We propose AnchorPrompt, an efficient adaptation method that keeps the model fr...

Feed lens

eval

Read the original at arxiv.org →Open in live feed Read that day’s brief

AnchorPrompt: Self-Distilled Soft Prompts for Robust Audio-Language Models

Earlier in this thread 4 items

RAMP: Robust Adaptive Mixed-Precision Quantization for Edge CPU Vision Models

RATL: Learning from Retrieved Residuals for Robust Multivariate Time-Series Forecasting

SCoNE: Selective Context-aware Neuron Editing for Robust Retrieval-Augmented Generation

DSPrompt: Dynamic Soft Prompt Defense Against M-RAG Corruption