paper
Listening with Attention: Entropy-Guided Explainability for Transformer-Based Audio Models
paperactiveprovisional
listening-with-attention-entropy-guided-explainability-for-transformer-based-audio-models-624892ec·1 events·first seen 2d agoAliases: Listening with Attention: Entropy-Guided Explainability for Transformer-Based Audio Models
Co-occurring entities
More like this (12)
Beyond Fully Random Masking: Attention-Guided Denoising and Optimization for Diffusion Language ModelsGenerative Explainability for Next-Generation Networks: LLM-Augmented XAI with Mutual Feature InteractionsGenerative Explainability for Next-Generation Networks: LLM-Augmented XAI with Mutual Feature InteractionsLeveraging Audio-LLMs to Filter Speech-to-Speech Training DataLanguage Modeling Lossencoder-only language modelsDecomposing Factual Sycophancy in Language Models: How Size and Instruction Tuning Shape RobustnessBeyond task performance: Decoding bioacoustic embeddings with speech featuresTransformer Language ModelsLanguage Model FinetuningSpeaker Group Encoding in Self-supervised Speech Recognition ModelsAnatomy of Post-Training: Using Interpretability to Characterize Data and Shape the Learning Signal
Recent events (1)
LEAF-X: Entropy-guided explainability framework for transformer-based ASR models
Researchers introduce LEAF-X (Listening with Entropy-guided Attention for Faithful explainability), a model-intrinsic XAI framework for transformer-based automatic speech recognition systems like Whisper. The method combines entropy-guided attention weighting, multi-layer attention rollout, and optional causal ablations to produce sparse token-to-frame attributions. Evaluations show 32% improved faithfulness and 35-39% stronger locality/sparsity compared to perturbation-based explainers and raw attention maps, enabling more auditable ASR.