Imagen de Portada

Shadows in the Attention: Contextual Perturbation and Representation Drift in the Dynamics of Hallucination in LLMs

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Wei, Zeyu, Wang, Shuo, Rong, Xiaohui, Liu, Xuemin, Li, He
Formato:	Preprint
Publicado:	2025
Materias:	Computation and Language
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Detecting Contextual Hallucinations in LLMs with Frequency-Aware Attention
por: Qi, Siya, et al.
Publicado: (2026)

AggTruth: Contextual Hallucination Detection using Aggregated Attention Scores in LLMs
por: Matys, Piotr, et al.
Publicado: (2025)

Probabilistic Guarantees for Reducing Contextual Hallucinations in LLMs
por: Rautenberg, Nils, et al.
Publicado: (2026)

Gradient-guided Attention Map Editing: Towards Efficient Contextual Hallucination Mitigation
por: Wang, Yu, et al.
Publicado: (2025)

Decoding by Perturbation: Mitigating MLLM Hallucinations via Dynamic Textual Perturbation
por: Jia, Sihang, et al.
Publicado: (2026)

LLM-CAS: Dynamic Neuron Perturbation for Real-Time Hallucination Correction
por: Zhang, Jensen, et al.
Publicado: (2025)

Adversarial Text Generation with Dynamic Contextual Perturbation
por: Waghela, Hetvi, et al.
Publicado: (2025)

Structured Knowledge Representation through Contextual Pages for Retrieval-Augmented Generation
por: Li, Xinze, et al.
Publicado: (2026)

Hallucination Detection in LLMs with Topological Divergence on Attention Graphs
por: Bazarova, Alexandra, et al.
Publicado: (2025)

Why LLMs Hallucinate on Structured Knowledge: A Mechanistic Analysis of Reasoning over Linearized Representations
por: Li, Shanghao, et al.
Publicado: (2026)

Wisdom is Knowing What not to Say: Hallucination-Free LLMs Unlearning via Attention Shifting
por: Tan, Chenchen, et al.
Publicado: (2025)

Look Within, Why LLMs Hallucinate: A Causal Perspective
por: Li, He, et al.
Publicado: (2024)

Mitigating Object Hallucinations in MLLMs via Multi-Frequency Perturbations
por: Li, Shuo, et al.
Publicado: (2025)

Speech LLMs are Contextual Reasoning Transcribers
por: Deng, Keqi, et al.
Publicado: (2026)

Do LLMs Know What Is Private Internally? Probing and Steering Contextual Privacy Norms in Large Language Model Representations
por: Wang, Haoran, et al.
Publicado: (2026)

Shakespearean Sparks: The Dance of Hallucination and Creativity in LLMs' Decoding Layers
por: He, Zicong, et al.
Publicado: (2025)

Counterfactual Debating with Preset Stances for Hallucination Elimination of LLMs
por: Fang, Yi, et al.
Publicado: (2024)

DAMRO: Dive into the Attention Mechanism of LVLM to Reduce Object Hallucination
por: Gong, Xuan, et al.
Publicado: (2024)

Attention Consistency for LLMs Explanation
por: Lan, Tian, et al.
Publicado: (2025)

PFME: A Modular Approach for Fine-grained Hallucination Detection and Editing of Large Language Models
por: Deng, Kunquan, et al.
Publicado: (2024)

Hallucination Detection in LLMs Using Spectral Features of Attention Maps
por: Binkowski, Jakub, et al.
Publicado: (2025)

Alleviating Forgetfulness of Linear Attention by Hybrid Sparse Attention and Contextualized Learnable Token Eviction
por: He, Mutian, et al.
Publicado: (2025)

Dynamic Attention-Guided Context Decoding for Mitigating Context Faithfulness Hallucinations in Large Language Models
por: Huang, Yanwen, et al.
Publicado: (2025)

Beyond Homogeneous Attention: Memory-Efficient LLMs via Fourier-Approximated KV Cache
por: Liu, Xiaoran, et al.
Publicado: (2025)

Think More, Hallucinate Less: Mitigating Hallucinations via Dual Process of Fast and Slow Thinking
por: Cheng, Xiaoxue, et al.
Publicado: (2025)

Understanding the Ability of LLMs to Handle Character-Level Perturbation
por: Zhuo, Anyuan, et al.
Publicado: (2025)

Detecting Hallucinations in SpeechLLMs at Inference Time Using Attention Maps
por: Waldendorf, Jonas, et al.
Publicado: (2026)

Rowen: Adaptive Retrieval-Augmented Generation for Hallucination Mitigation in LLMs
por: Ding, Hanxing, et al.
Publicado: (2024)

Lookback Lens: Detecting and Mitigating Contextual Hallucinations in Large Language Models Using Only Attention Maps
por: Chuang, Yung-Sung, et al.
Publicado: (2024)

Instruction Tuning and CoT Prompting for Contextual Medical QA with LLMs
por: Le, Chenqian, et al.
Publicado: (2025)

Can LLMs Detect Their Own Hallucinations?
por: Kadotani, Sora, et al.
Publicado: (2025)

Shared Imagination: LLMs Hallucinate Alike
por: Zhou, Yilun, et al.
Publicado: (2024)

OPDAI at SemEval-2024 Task 6: Small LLMs can Accelerate Hallucination Detection with Weakly Supervised Data
por: Wei, Chengcheng, et al.
Publicado: (2024)

Beyond Logit Lens: Contextual Embeddings for Robust Hallucination Detection & Grounding in VLMs
por: Phukan, Anirudh, et al.
Publicado: (2024)

Benchmarking Contextual and Paralinguistic Reasoning in Speech-LLMs: A Case Study with In-the-Wild Data
por: Wang, Qiongqiong, et al.
Publicado: (2025)

Toxic HallucinAItions: Perturbing Prompts and Tracing LLM Circuits
por: Shimgekar, Soorya Ram, et al.
Publicado: (2026)

Continuously Steering LLMs Sensitivity to Contextual Knowledge with Proxy Models
por: Wang, Yilin, et al.
Publicado: (2025)

Mitigating Object Hallucination via Concentric Causal Attention
por: Xing, Yun, et al.
Publicado: (2024)

KDCM: Reducing Hallucination in LLMs through Explicit Reasoning Structures
por: Hao, Jinbo, et al.
Publicado: (2026)

Med-HEAL: Analyzing and Mitigating Hallucinations in Medical LLMs with Hallucination-Aware In-Context Learning
por: Liao, Yiming, et al.
Publicado: (2026)