Image de couverture de livre

The Attribution Blind Spot: Detecting When Language Models Rely on Memory Rather Than Retrieved Context

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Yu, Zhe, Xing, Wenpeng, Wei, Yunzhao, Yang, Bo, Ye, Chen, Li, Gaolei, Han, Meng
Format:	Preprint
Publié:	2026
Sujets:	Artificial Intelligence
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

LatentAudit: Real-Time White-Box Faithfulness Monitoring for Retrieval-Augmented Generation with Verifiable Deployment
par: Yu, Zhe, et autres
Publié: (2026)

Composition Collapse: Stable Factual Knowledge Does Not Imply Compositional Reasoning
par: Yu, Zhe, et autres
Publié: (2026)

From Retinal Evidence to Safe Decisions: RETINA-SAFE and ECRT for Hallucination Risk Triage in Medical LLMs
par: Yu, Zhe, et autres
Publié: (2026)

Cordon-MAS: Defending RAG against Knowledge Poisoning via Information-Flow Control
par: Yu, Zhe, et autres
Publié: (2026)

Anti‐Tumour Immunity Relies on Targeting Tissue Homeostasis Through Monocyte‐Driven Responses Rather Than Direct Tumour Cytotoxicity
par: Nicholas Koelsch, et autres
Publié: (2025)

KinGuard: Hierarchical Kinship-Aware Fingerprinting to Defend Against Large Language Model Stealing
par: Xu, Zhenhua, et autres
Publié: (2026)

Blinded by Generated Contexts: How Language Models Merge Generated and Retrieved Contexts When Knowledge Conflicts?
par: Tan, Hexiang, et autres
Publié: (2024)

When Do Individuals Believe in Themselves Rather Than in Artificial Intelligence? Insights from Longitudinal Investigations in Corporate Credit‐Rating Contexts
par: Kyootai Lee, et autres
Publié: (2025)

EverTracer: Hunting Stolen Large Language Models via Stealthy and Robust Probabilistic Fingerprint
par: Xu, Zhenhua, et autres
Publié: (2025)

Why Is There I Rather Than It?
par: Kudelska, Marta
Publié: (2022)

To Rely or Not to Rely? Evaluating Interventions for Appropriate Reliance on Large Language Models
par: Bo, Jessica Y., et autres
Publié: (2024)

AI in the Clinical Radiology and Radiation Oncology Assessed Curricula: When Rather Than If
par: Daniel E. Roos, et autres
Publié: (2025)

HGMF: A Hierarchical Gaussian Mixture Framework for Scalable Tool Invocation within the Model Context Protocol
par: Xing, Wenpeng, et autres
Publié: (2025)

Why Is There Something Rather Than Nothing?
par: Karakülah, Bülent
Publié: (2026)

Observing Interaction Rather Than Interfaces
par: Rivière, Guillaume
Publié: (2025)

TeacherLM: Teaching to Fish Rather Than Giving the Fish, Language Modeling Likewise
par: He, Nan, et autres
Publié: (2023)

CoDe-NeRF: Neural Rendering via Dynamic Coefficient Decomposition
par: Xing, Wenpeng, et autres
Publié: (2025)

Learning about Learning Rather Than about Teaching.
par: Markless, Sharon
Publié: (2002)

Social Equality Is Rather Threatening Than Normative to Authoritarians
par: Márton Hadarics, et autres
Publié: (2025)

Linguistic Blind Spots of Large Language Models
par: Cheng, Jiali, et autres
Publié: (2025)

Temporal Blind Spots in Large Language Models
par: Wallat, Jonas, et autres
Publié: (2024)

NeuRel-Attack: Neuron Relearning for Safety Disalignment in Large Language Models
par: Zhou, Yi, et autres
Publié: (2025)

Blind OFDM-ISAC Relying on Asymmetric Modem Constellations
par: Pu, Henglin, et autres
Publié: (2026)

Incorporating Rather Than Eliminating: Achieving Fairness for Skin Disease Diagnosis Through Group-Specific Expert
par: Xu, Gelei, et autres
Publié: (2025)

Structured Memory Mechanisms for Stable Context Representation in Large Language Models
par: Xing, Yue, et autres
Publié: (2025)

With Argus Eyes: Assessing Retrieval Gaps via Uncertainty Scoring to Detect and Remedy Retrieval Blind Spots
par: Taghavi, Zeinab Sadat, et autres
Publié: (2026)

Simple Linguistic Inferences of Large Language Models (LLMs): Blind Spots and Blinds
par: Basmov, Victoria, et autres
Publié: (2023)

Fluent but Unfeeling: The Emotional Blind Spots of Language Models
par: Shu, Bangzhao, et autres
Publié: (2025)

PREE: Towards Harmless and Adaptive Fingerprint Editing in Large Language Models via Knowledge Prefix Enhancement
par: Yue, Xubin, et autres
Publié: (2025)

Blind Spot Navigation in Large Language Model Reasoning with Thought Space Explorer
par: Zhang, Jinghan, et autres
Publié: (2024)

Effects of Statistical Learning on Stimulus Detection Speed Reflect Attention Rather Than Representational Changes
par: Yoel Sheinenson, et autres
Publié: (2026)

Blind Spots
par: Drucker, Johanna
Publié: (2009)

Blind Spot Navigation: Evolutionary Discovery of Sensitive Semantic Concepts for LVLMs
par: Pan, Zihao, et autres
Publié: (2025)

Selective Annotation via Data Allocation: These Data Should Be Triaged to Experts for Annotation Rather Than the Model
par: Huang, Chen, et autres
Publié: (2024)

WhenLoss: Diagnosing Write and Retrieval Bottlenecks in Long-Context Memory Systems
par: Yu, Jiangnan, et autres
Publié: (2026)

Memory-Augmented Large Language Models: External Memory, Retrieval Mechanisms, and Long-Context Processing
par: Junjie, Yang
Publié: (2026)

Illuminating Blind Spots of Language Models with Targeted Agent-in-the-Loop Synthetic Data
par: Lippmann, Philip, et autres
Publié: (2024)

Verification Blind Spots: When Branchless Code Defeats Static Analysis
par: Musselman, Blake
Publié: (2026)

Incentivize Early Renewals Rather Than Train Members to Pay Late
Publié: (2024)

SproutBench: A Benchmark for Safe and Ethical Large Language Models for Youth
par: Xing, Wenpeng, et autres
Publié: (2025)