# Do LLMs Really Know What They Don't Know? Internal States Mainly Reflect Knowledge Recall Rather Than Truthfulness

arXiv: https://arxiv.org/abs/2510.09033 · HTML: https://mlsanity.com/paper/2510.09033/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- Builds on 25 of its 44 references that are in the map.
- First posted on arXiv 10 Oct 2025.

## Builds on (most cited first, 25 of 25)

- [QLoRA: Efficient Finetuning of Quantized LLMs](https://mlsanity.com/paper/2305.14314/) (2023)
- [TruthfulQA: Measuring How Models Mimic Human Falsehoods](https://mlsanity.com/paper/2109.07958/) (2021)
- [HuggingFace's Transformers: State-of-the-art Natural Language Processing](https://mlsanity.com/paper/1910.03771/) (2019)
- [Inference-Time Intervention: Eliciting Truthful Answers from a Language Model](https://mlsanity.com/paper/2306.03341/) (2023)
- [FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generation](https://mlsanity.com/paper/2305.14251/) (2023)
- [When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memories](https://mlsanity.com/paper/2212.10511/) (2022)
- [Can LLMs Express Their Uncertainty? An Empirical Evaluation of Confidence Elicitation in LLMs](https://mlsanity.com/paper/2306.13063/) (2023)
- [Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback](https://mlsanity.com/paper/2305.14975/) (2023)
- [Teaching Models to Express Their Uncertainty in Words](https://mlsanity.com/paper/2205.14334/) (2022)
- [Dissecting Recall of Factual Associations in Auto-Regressive Language Models](https://mlsanity.com/paper/2304.14767/) (2023)
- [LLMs Know More Than They Show: On the Intrinsic Representation of LLM Hallucinations](https://mlsanity.com/paper/2410.02707/) (2024)
- [Do Large Language Models Know What They Don't Know?](https://mlsanity.com/paper/2305.18153/) (2023)
- [A Stitch in Time Saves Nine: Detecting and Mitigating Hallucinations of LLMs by Validating Low-Confidence Generation](https://mlsanity.com/paper/2307.03987/) (2023)
- [Simple Entity-Centric Questions Challenge Dense Retrievers](https://mlsanity.com/paper/2109.08535/) (2021)
- [SAC3: Reliable Hallucination Detection in Black-Box Language Models via Semantic-aware Cross-check Consistency](https://mlsanity.com/paper/2311.01740/) (2023)
- [LLM Internal States Reveal Hallucination Risk Faced With a Query](https://mlsanity.com/paper/2407.03282/) (2024)
- [Attention Satisfies: A Constraint-Satisfaction Lens on Factual Errors of Language Models](https://mlsanity.com/paper/2309.15098/) (2023)
- [When Do LLMs Need Retrieval Augmentation? Mitigating LLMs' Overconfidence Helps Retrieval Augmentation](https://mlsanity.com/paper/2402.11457/) (2024)
- [Knowing What LLMs DO NOT Know: A Simple Yet Effective Self-Detection Method](https://mlsanity.com/paper/2310.17918/) (2023)
- [Estimating Knowledge in Large Language Models Without Generating a Single Token](https://mlsanity.com/paper/2406.12673/) (2024)
- [Reasoning about Uncertainty: Do Reasoning Models Know When They Don't Know?](https://mlsanity.com/paper/2506.18183/) (2025)
- [Towards Fully Exploiting LLM Internal States to Enhance Knowledge Boundary Perception](https://mlsanity.com/paper/2502.11677/) (2025)
- [Narrowing the Knowledge Evaluation Gap: Open-Domain Question Answering with Multi-Granularity Answers](https://mlsanity.com/paper/2401.04695/) (2024)
- [Can LMs Generalize to Future Data? An Empirical Analysis on Text Summarization](https://mlsanity.com/paper/2305.01951/) (2023)
- [Analyzing LLMs' Knowledge Boundary Cognition Across Languages Through the Lens of Internal Representations](https://mlsanity.com/paper/2504.13816/) (2025)

## Built on by (most cited first, 10 of 10)

- [The strength of clinical evidence is recoverable from language model representations but not from their stated grades](https://mlsanity.com/paper/2606.29034/) (2026)
- [Diagnosing Correctness Probes under Self-Judgement Confounding](https://mlsanity.com/paper/2607.16799/) (2026)
- [MobileMem: Learning from a Year of Mobile Experiences](https://mlsanity.com/paper/2608.13606/) (2026)
- [Temporal Multi-Signal Fusion for Token-Level Hallucination Detection](https://mlsanity.com/paper/2608.18115/) (2026)
- [Legible Failures: Detecting and Repairing In-Context Binding Errors](https://mlsanity.com/paper/2609.11216/) (2026)
- [Decodable In-Context State and Model Output Across Training](https://mlsanity.com/paper/2609.31401/) (2026)
- [Knowing Is Not Choosing: What Explicit Verification Adds Beyond Generative Preference](https://mlsanity.com/paper/2609.33142/) (2026)
- [Beyond the Shadows of Plato's Cave: Evaluating False Memory in Autonomous Agents via Counterfactual Reasoning](https://mlsanity.com/paper/2609.39473/) (2026)
- [The Geometry of Knowledge Accessibility in Large Language Models](https://mlsanity.com/paper/2610.03052/) (2026)
- [Visual Dependence Does Not Imply Answerability: A Systematic Analysis of Answerability Signals in Frozen InstructBLIP](https://mlsanity.com/paper/2610.17902/) (2026)

## Abstract (arXiv)

Recent work suggests that LLMs "know what they don't know", positing that hallucinated and factually correct outputs arise from distinct internal processes and can therefore be distinguished using internal signals. However, hallucinations have multifaceted causes: beyond simple knowledge gaps, they can emerge from training incentives that encourage models to exploit statistical shortcuts or spurious associations learned during pretraining. In this paper, we argue that when LLMs rely on such learned associations to produce hallucinations, their internal processes are mechanistically similar to those of factual recall, as both stem from strong statistical correlations encoded in the model's parameters. To verify this, we propose a novel taxonomy categorizing hallucinations into Unassociated Hallucinations (UHs), where outputs lack parametric grounding, and Associated Hallucinations (AHs), which are driven by spurious associations. Through mechanistic analysis, we compare their computational processes and hidden-state geometries with factually correct outputs. Our results show that hidden states primarily reflect whether the model is recalling parametric knowledge rather than the truthfulness of the output itself. Consequently, AHs exhibit hidden-state geometries that largely overlap with factual outputs, rendering standard detection methods ineffective. In contrast, UHs exhibit distinctive, clustered representations that facilitate reliable detection.