paper-with-me

홈 › Papers

From Dispersion to Attraction: Spectral Dynamics of Hallucination Across Whisper Model Scales

2026-03-31 · Ivan Viakhirev, Kirill Borodin, Grach Mkrtchian arxiv

Hallucinations in large ASR models present a critical safety risk. In this work, we propose the \textit{Spectral Sensitivity Theorem}, which predicts a phase transition in deep networks from a dispersive regime (signal decay) to an attractor regime (rank-1 collapse) governed by layer-wise gain and alignment. We validate this theory by analyzing the eigenspectra of activation graphs in Whisper models (Tiny to Large-v3-Turbo) under adversarial stress. Our results confirm the theoretical prediction: intermediate models exhibit \textit{Structural Disintegration} (Regime I), characterized by a $13.4\%$ collapse in Cross-Attention rank. Conversely, large models enter a \textit{Compression-Seeking Attractor} state (Regime II), where Self-Attention actively compresses rank ($-2.34\%$) and hardens the spectral slope, decoupling the model from acoustic evidence.

📄 PDF Abstract BibTeX arXiv:2604.08591

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

D$^2$HScore: Reasoning-Aware Hallucination Detection via Semantic Breadth and Depth Analysis in LLMs

2025-09-15 · Yue Ding, Xiaofang Zhu, Tianze Xia, Junfei Wu 외 arxiv

Although large Language Models (LLMs) have achieved remarkable success, their practical application is often hindered by the generation of non-factual content, which is called "hallucination". Ensuring the reliability of…

Entropy and Attention Dynamics in Small Language Models: A Trace-Level Structural Analysis on the TruthfulQA Benchmark

2026-04-04 · Adeyemi Adeseye, Aisvarya Adeseye, Hannu Tenhunen, Jouni Isoaho arxiv

Small language models (SLMs) have been increasingly deployed in edge devices and other resource-constrained settings. However, these models make confident mispredictions and produce unstable output, making them risky for…

NerVE: Nonlinear Eigenspectrum Dynamics in LLM Feed-Forward Networks

2026-03-06 · Nandan Kumar Jha, Brandon Reagen arxiv

We introduce NerVE, a unified eigenspectral framework for understanding how feed-forward networks (FFNs) in large language models (LLMs) organize and regulate information flow in high-dimensional latent space. Despite FF…

Grounding the Ungrounded: A Spectral-Graph Framework for Quantifying Hallucinations in Multimodal LLMs

2025-08-26 · Supratik Sarkar, Swagatam Das arxiv

Hallucinations in LLMs--especially in multimodal settings--undermine reliability. We present a rigorous information-geometric framework, grounded in diffusion dynamics, to quantify hallucinations in MLLMs where model out…

HICD: Hallucination-Inducing via Attention Dispersion for Contrastive Decoding to Mitigate Hallucinations in Large Language Models

2025-03-17 · Xinyan Jiang, Hang Ye, Yongxin Zhu, Xiaoying Zheng 외

Large Language Models (LLMs) often generate hallucinations, producing outputs that are contextually inaccurate or factually incorrect. We introduce HICD, a novel method designed to induce hallucinations for contrastive d…

HallucinationQuestion AnsweringReading Comprehension