Hallucination Across the Reasoning Lifecycle: Interface Visibility, Causal Evidence, and Release Control in Large Reasoning Models

arXiv:2610.05472 · cs.AI · Submitted 2026-10-04 · Read on arXiv

cs.AI

Submitted: 2026-10-04

Updated: 2026-10-04

Comments: 35 pages, 8 figures, 9 tables. Electronic supplementary materials S1-S7 are included as ancillary files

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Hallucination — 1× in this paper · explained in 4 episodes

Sources

• Large Language Models Hallucination: A Comprehensive Survey
• How Much Does a Reasoning Summary Reveal? An Observability Ladder for Large Language Models
• Do I Really Know? Learning Factual Self-Verification for Hallucination Reduction
• The Periodic Table of LLM Reasoning: A Structured Survey of Reasoning Paradigms, Methods, and Failure Modes
• Two Regimes of Chain-of-Thought Unfaithfulness: Metric-Based Detection Fails Where Models Are Wrong
• Chain-of-Thought Reasoning In The Wild Is Not Always Faithful
• LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding
• Information-seeking failures of large language models in agentic clinical reasoning
• HalluSAE: Detecting Hallucinations in Large Language Models via Sparse Auto-Encoders
• INSIDE: LLMs' Internal States Retain the Power of Hallucination Detection
• HaluMem: Evaluating Hallucinations in Memory Systems of Agents
• Reflect-R1: Evidence-Driven Reflection for Self-Correction in Long Video Understanding
• Learning to Reason for Factuality
• Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs
• The FACTS Leaderboard: A Comprehensive Benchmark for Large Language Model Factuality
• Chain-of-Thought Prompting Obscures Hallucination Cues in Large Language Models: An Empirical Evaluation
• Think More, Hallucinate Less: Mitigating Hallucinations via Dual Process of Fast and Slow Thinking
• FacTool: Factuality Detection in Generative AI -- A Tool Augmented Framework for Multi-Task and Multi-Domain Scenarios
• Are DeepSeek R1 And Other Reasoning Models More Faithful?
• DoLa: Decoding by Contrasting Layers Improves Factuality in Large Language Models

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• Agents' Last Exam
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems