Imagen de Portada

How Transformers Reject Wrong Answers: Rotational Dynamics of Factual Constraint Processing

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autor principal:	Marín, Javier
Formato:	Preprint
Publicado:	2026
Materias:	Computation and Language Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Varying Shades of Wrong: Aligning LLMs with Wrong Answers Only
por: Yao, Jihan, et al.
Publicado: (2024)

Empirical Characterization of Temporal Constraint Processing in LLMs
por: Marín, Javier
Publicado: (2025)

Promote, Suppress, Iterate: How Language Models Answer One-to-Many Factual Queries
por: Yan, Tianyi Lorena, et al.
Publicado: (2025)

The Curious Case of Factual (Mis)Alignment between LLMs' Short- and Long-Form Answers
por: Islam, Saad Obaid ul, et al.
Publicado: (2025)

Evaluating Reliability Asymmetries in Chinese Factual Search and AI Answers
por: Liu, Geng, et al.
Publicado: (2025)

CounterRefine: Answer-Conditioned Counterevidence Retrieval for Inference-Time Knowledge Repair in Factual Question Answering
por: Huang, Tianyi, et al.
Publicado: (2026)

Answer, Assemble, Ace: Understanding How LMs Answer Multiple Choice Questions
por: Wiegreffe, Sarah, et al.
Publicado: (2024)

When Verification Fails: How Compositionally Infeasible Claims Escape Rejection
por: Liu, Muxin, et al.
Publicado: (2026)

Trustworthy Reasoning: Evaluating and Enhancing Factual Accuracy in LLM Intermediate Thought Processes
por: Jiao, Rui, et al.
Publicado: (2025)

Factuality on Demand: Controlling the Factuality-Informativeness Trade-off in Text Generation
por: Gong, Ziwei, et al.
Publicado: (2026)

Do Automatic Factuality Metrics Measure Factuality? A Critical Evaluation
por: Ramprasad, Sanjana, et al.
Publicado: (2024)

DyKnow: Dynamically Verifying Time-Sensitive Factual Knowledge in LLMs
por: Mousavi, Seyed Mahed, et al.
Publicado: (2024)

No Clustering, No Routing: How Transformers Actually Process Rare Tokens
por: Liu, Jing
Publicado: (2025)

Attention Satisfies: A Constraint-Satisfaction Lens on Factual Errors of Language Models
por: Yuksekgonul, Mert, et al.
Publicado: (2023)

How Does Response Length Affect Long-Form Factuality
por: Zhao, James Xu, et al.
Publicado: (2025)

Correct Chains, Wrong Answers: Dissociating Reasoning from Output in LLM Logic
por: Rao, Abinav, et al.
Publicado: (2026)

Correct Answers from Sound Reasoning: Verifiable Process Supervision for Language Models
por: Kim, Kyuyoung, et al.
Publicado: (2026)

ADEPT: Adaptive Dynamic Early-Exit Process for Transformers
por: Yoo, Sangmin, et al.
Publicado: (2026)

Is Factuality Enhancement a Free Lunch For LLMs? Better Factuality Can Lead to Worse Context-Faithfulness
por: Bi, Baolong, et al.
Publicado: (2024)

Correct after Answer: Enhancing Multi-Span Question Answering with Post-Processing Method
por: Lin, Jiayi, et al.
Publicado: (2024)

Not Wrong, But Untrue: LLM Overconfidence in Document-Based Queries
por: Hagar, Nick, et al.
Publicado: (2025)

What's Wrong? Refining Meeting Summaries with LLM Feedback
por: Kirstein, Frederic, et al.
Publicado: (2024)

Interpreting Key Mechanisms of Factual Recall in Transformer-Based Language Models
por: Lv, Ang, et al.
Publicado: (2024)

A Geometric Taxonomy of Hallucinations in LLMs
por: Marín, Javier
Publicado: (2026)

Multiple Choice Questions: Reasoning Makes Large Language Models (LLMs) More Self-Confident, Especially When They are Wrong
por: Fu, Tairan, et al.
Publicado: (2025)

How do you know that? Teaching Generative Language Models to Reference Answers to Biomedical Questions
por: Bašaragin, Bojana, et al.
Publicado: (2024)

Revisiting Self-Consistency from Dynamic Distributional Alignment Perspective on Answer Aggregation
por: Li, Yiwei, et al.
Publicado: (2025)

On Early Detection of Hallucinations in Factual Question Answering
por: Snyder, Ben, et al.
Publicado: (2023)

Generating Benchmarks for Factuality Evaluation of Language Models
por: Muhlgay, Dor, et al.
Publicado: (2023)

Language Models' Factuality Depends on the Language of Inquiry
por: Aggarwal, Tushar, et al.
Publicado: (2025)

From Confidence to Collapse in LLM Factual Robustness
por: Fastowski, Alina, et al.
Publicado: (2025)

Factuality of Large Language Models: A Survey
por: Wang, Yuxia, et al.
Publicado: (2024)

Sandwich Reasoning: An Answer-Reasoning-Answer Approach for Low-Latency Query Correction
por: Zhang, Chen, et al.
Publicado: (2026)

When the Majority is Wrong: Modeling Annotator Disagreement for Subjective Tasks
por: Fleisig, Eve, et al.
Publicado: (2023)

FactReasoner: A Probabilistic Approach to Long-Form Factuality Assessment for Large Language Models
por: Marinescu, Radu, et al.
Publicado: (2025)

No Answer Needed: Predicting LLM Answer Accuracy from Question-Only Linear Probes
por: Cencerrado, Iván Vicente Moreno, et al.
Publicado: (2025)

Permutation-Consensus Listwise Judging for Robust Factuality Evaluation
por: Huang, Tianyi, et al.
Publicado: (2026)

InFact: Informativeness Alignment for Improved LLM Factuality
por: Cohen, Roi, et al.
Publicado: (2025)

Real-time Factuality Assessment from Adversarial Feedback
por: Chen, Sanxing, et al.
Publicado: (2024)

Aligning Knowledge Graphs and Language Models for Factual Accuracy
por: Nishat, Nur A Zarin, et al.
Publicado: (2025)