Imagen de Portada

Why and How LLMs Hallucinate: Connecting the Dots with Subsequence Associations

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Sun, Yiyou, Gai, Yu, Chen, Lijie, Ravichander, Abhilasha, Choi, Yejin, Song, Dawn
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

HALoGEN: Fantastic LLM Hallucinations and Where to Find Them
por: Ravichander, Abhilasha, et al.
Publicado: (2025)

Artifacts or Abduction: How Do LLMs Answer Multiple-Choice Questions Without the Question?
por: Balepur, Nishant, et al.
Publicado: (2024)

RESTOR: Knowledge Recovery in Machine Unlearning
por: Rezaei, Keivan, et al.
Publicado: (2024)

WildHallucinations: Evaluating Long-form Factuality in LLMs with Real-World Entity Queries
por: Zhao, Wenting, et al.
Publicado: (2024)

What Has Been Lost with Synthetic Evaluation?
por: Gill, Alexander, et al.
Publicado: (2025)

RL Grokking Recipe: How Does RL Unlock and Transfer New Algorithms in LLMs?
por: Sun, Yiyou, et al.
Publicado: (2025)

WildBench: Benchmarking LLMs with Challenging Tasks from Real Users in the Wild
por: Lin, Bill Yuchen, et al.
Publicado: (2024)

The Curious Case of Factuality Finetuning: Models' Internal Beliefs Can Improve Factuality
por: Newman, Benjamin, et al.
Publicado: (2025)

Information-Guided Identification of Training Data Imprint in (Proprietary) Large Language Models
por: Ravichander, Abhilasha, et al.
Publicado: (2025)

Agent Lumos: Unified and Modular Training for Open-Source Language Agents
por: Yin, Da, et al.
Publicado: (2023)

The Surprising Effectiveness of Membership Inference with Simple N-Gram Coverage
por: Hallinan, Skyler, et al.
Publicado: (2025)

OMEGA: Can LLMs Reason Outside the Box in Math? Evaluating Exploratory, Compositional, and Transformative Generalization
por: Sun, Yiyou, et al.
Publicado: (2025)

Can LLMs Ask Good Questions?
por: Zhang, Yueheng, et al.
Publicado: (2025)

Fractional Rotation, Full Potential? Investigating Performance and Convergence of Partial RoPE
por: Khan, Mohammad Aflah, et al.
Publicado: (2026)

MacGyver: Are Large Language Models Creative Problem Solvers?
por: Tian, Yufei, et al.
Publicado: (2023)

Climbing the Ladder of Reasoning: What LLMs Can-and Still Can't-Solve after SFT?
por: Sun, Yiyou, et al.
Publicado: (2025)

KnowHalu: Hallucination Detection via Multi-Form Knowledge Based Factual Checking
por: Zhang, Jiawei, et al.
Publicado: (2024)

Reverse Question Answering: Can an LLM Write a Question so Hard (or Bad) that it Can't Answer?
por: Balepur, Nishant, et al.
Publicado: (2024)

Strategy Executability in Mathematical Reasoning: Leveraging Human-Model Differences for Effective Guidance
por: Liang, Weida, et al.
Publicado: (2026)

How much reliable is ChatGPT's prediction on Information Extraction under Input Perturbations?
por: Mondal, Ishani, et al.
Publicado: (2024)

CodaRAG: Connecting the Dots with Associativity Inspired by Complementary Learning
por: Li, Cheng-Yen, et al.
Publicado: (2026)

Can LLMs Reason with Rules? Logic Scaffolding for Stress-Testing and Improving LLMs
por: Wang, Siyuan, et al.
Publicado: (2024)

H-Neurons: On the Existence, Impact, and Origin of Hallucination-Associated Neurons in LLMs
por: Gao, Cheng, 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)

Being Kind Isn't Always Being Safe: Diagnosing Affective Hallucination in LLMs
por: Kim, Sewon, et al.
Publicado: (2025)

CodeHalu: Investigating Code Hallucinations in LLMs via Execution-based Verification
por: Tian, Yuchen, et al.
Publicado: (2024)

DailyDilemmas: Revealing Value Preferences of LLMs with Quandaries of Daily Life
por: Chiu, Yu Ying, et al.
Publicado: (2024)

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

In Agents We Trust, but Who Do Agents Trust? Latent Source Preferences Steer LLM Generations
por: Khan, Mohammad Aflah, et al.
Publicado: (2026)

DFA-RAG: Conversational Semantic Router for Large Language Model with Definite Finite Automaton
por: Sun, Yiyou, et al.
Publicado: (2024)

Why Language Models Hallucinate
por: Kalai, Adam Tauman, et al.
Publicado: (2025)

The Art of Saying No: Contextual Noncompliance in Language Models
por: Brahman, Faeze, et al.
Publicado: (2024)

Why Fine-Tuning Encourages Hallucinations and How to Fix It
por: Kaplan, Guy, et al.
Publicado: (2026)

Unsafer in Many Turns: Benchmarking and Defending Multi-Turn Safety Risks in Tool-Using Agents
por: Li, Xu, et al.
Publicado: (2026)

From Single to Multi: How LLMs Hallucinate in Multi-Document Summarization
por: Belem, Catarina G., et al.
Publicado: (2024)

Connecting the Dots: LLMs can Infer and Verbalize Latent Structure from Disparate Training Data
por: Treutlein, Johannes, et al.
Publicado: (2024)

Why LLMs Cannot Think and How to Fix It
por: Jahrens, Marius, et al.
Publicado: (2025)

How Much Do LLMs Hallucinate across Languages? On Realistic Multilingual Estimation of LLM Hallucination
por: Islam, Saad Obaid ul, et al.
Publicado: (2025)

Understanding How Value Neurons Shape the Generation of Specified Values in LLMs
por: Su, Yi, et al.
Publicado: (2025)

LLM Internal States Reveal Hallucination Risk Faced With a Query
por: Ji, Ziwei, et al.
Publicado: (2024)