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

arXiv: https://arxiv.org/abs/2504.12691 · HTML: https://mlsanity.com/paper/2504.12691/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- Builds on 70 of its 79 references that are in the map.
- First posted on arXiv 17 Apr 2025.

## Builds on (most cited first, 30 of 70)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [A Unified Approach to Interpreting Model Predictions](https://mlsanity.com/paper/1705.07874/) (2017)
- ["Why Should I Trust You?": Explaining the Predictions of Any Classifier](https://mlsanity.com/paper/1602.04938/) (2016)
- [Neural Machine Translation by Jointly Learning to Align and Translate](https://mlsanity.com/paper/1409.0473/) (2014)
- [Axiomatic Attribution for Deep Networks](https://mlsanity.com/paper/1703.01365/) (2017)
- [WebArena: A Realistic Web Environment for Building Autonomous Agents](https://mlsanity.com/paper/2307.13854/) (2023)
- [Shortcut Learning in Deep Neural Networks](https://mlsanity.com/paper/2004.07780/) (2020)
- [Sanity Checks for Saliency Maps](https://mlsanity.com/paper/1810.03292/) (2018)
- [Beyond Accuracy: Behavioral Testing of NLP models with CheckList](https://mlsanity.com/paper/2005.04118/) (2020)
- [Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks](https://mlsanity.com/paper/1506.03099/) (2015)
- [The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora with Web Data, and Web Data Only](https://mlsanity.com/paper/2306.01116/) (2023)
- [Transformers learn in-context by gradient descent](https://mlsanity.com/paper/2212.07677/) (2022)
- [Siren's Song in the AI Ocean: A Survey on Hallucination in Large Language Models](https://mlsanity.com/paper/2309.01219/) (2023)
- [OLMo: Accelerating the Science of Language Models](https://mlsanity.com/paper/2402.00838/) (2024)
- [ERASER: A Benchmark to Evaluate Rationalized NLP Models](https://mlsanity.com/paper/1911.03429/) (2019)
- [Dolma: an Open Corpus of Three Trillion Tokens for Language Model Pretraining Research](https://mlsanity.com/paper/2402.00159/) (2024)
- [The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"](https://mlsanity.com/paper/2309.12288/) (2023)
- [Trained Transformers Learn Linear Models In-Context](https://mlsanity.com/paper/2306.09927/) (2023)
- [Android in the Wild: A Large-Scale Dataset for Android Device Control](https://mlsanity.com/paper/2307.10088/) (2023)
- [How Language Model Hallucinations Can Snowball](https://mlsanity.com/paper/2305.13534/) (2023)
- [Rationalizing Neural Predictions](https://mlsanity.com/paper/1606.04155/) (2016)
- [Transformers learn to implement preconditioned gradient descent for in-context learning](https://mlsanity.com/paper/2306.00297/) (2023)
- [Do Large Language Models Know What They Don't Know?](https://mlsanity.com/paper/2305.18153/) (2023)
- [Explaining by Removing: A Unified Framework for Model Explanation](https://mlsanity.com/paper/2011.14878/) (2020)
- [Probing Neural Network Comprehension of Natural Language Arguments](https://mlsanity.com/paper/1907.07355/) (2019)
- [Factuality Enhanced Language Models for Open-Ended Text Generation](https://mlsanity.com/paper/2206.04624/) (2022)
- [Visualizing and Understanding Neural Models in NLP](https://mlsanity.com/paper/1506.01066/) (2015)
- [Simple synthetic data reduces sycophancy in large language models](https://mlsanity.com/paper/2308.03958/) (2023)
- [Embers of Autoregression: Understanding Large Language Models Through the Problem They are Trained to Solve](https://mlsanity.com/paper/2309.13638/) (2023)

## Built on by (most cited first, 15 of 15)

- [Why Language Models Hallucinate](https://mlsanity.com/paper/2509.04664/) (2025)
- [CounselBench: A Large-Scale Expert Evaluation and Adversarial Benchmarking of Large Language Models in Mental Health Question Answering](https://mlsanity.com/paper/2506.08584/) (2025)
- [Invisible Tokens, Visible Bills: The Urgent Need to Audit Hidden Operations in Opaque LLM Services](https://mlsanity.com/paper/2505.18471/) (2025)
- [Honesty over Accuracy: Trustworthy Language Models through Reinforced Hesitation](https://mlsanity.com/paper/2511.11500/) (2025)
- [When Bias Pretends to Be Truth: How Spurious Correlations Undermine Hallucination Detection in LLMs](https://mlsanity.com/paper/2511.07318/) (2025)
- [Explainability of Large Language Models: Opportunities and Challenges toward Generating Trustworthy Explanations](https://mlsanity.com/paper/2510.17256/) (2025)
- [Hallucination Basins: A Dynamic Framework for Understanding and Controlling LLM Hallucinations](https://mlsanity.com/paper/2604.04743/) (2026)
- [The Piggyback Hypothesis of Generalization: Explaining and Mitigating Emergent Misalignment](https://mlsanity.com/paper/2606.06667/) (2026)
- [Actionable Hallucination Detection: Translating Latent Uncertainty into Agentic Critique](https://mlsanity.com/paper/2608.10430/) (2026)
- [The Truth Stays in the Family: Enhancing Contextual Grounding via Inherited Truthful Heads in Model Lineages](https://mlsanity.com/paper/2606.15821/) (2026)
- [Understanding Why Language Models Hallucinate: Testing Reasoning Against Priors](https://mlsanity.com/paper/2607.00447/) (2026)
- [When Guessing is Rewarded: Rethinking Language Model Evaluation with Distributional Uncertainty Scoring](https://mlsanity.com/paper/2510.04302/) (2025)
- [SoftMatcha 2: A Fast and Soft Pattern Matcher for Trillion-Scale Corpora](https://mlsanity.com/paper/2602.10908/) (2026)
- [When the Wrong Key Wins: Understanding and Detecting Hallucinations in LLMs](https://mlsanity.com/paper/2609.15106/) (2026)
- [Tracing and Taming Hallucinated Terms via Subsequence-Association Filtering and Breadth-Aware Selection for LLM Query Expansion](https://mlsanity.com/paper/2610.13810/) (2026)

## Abstract (arXiv)

Large language models (LLMs) frequently generate hallucinations-content that deviates from factual accuracy or provided context-posing challenges for diagnosis due to the complex interplay of underlying causes. This paper introduces a subsequence association framework to systematically trace and understand hallucinations. Our key insight is that hallucinations arise when dominant hallucinatory associations outweigh faithful ones. Through theoretical and empirical analyses, we demonstrate that decoder-only transformers effectively function as subsequence embedding models, with linear layers encoding input-output associations. We propose a tracing algorithm that identifies causal subsequences by analyzing hallucination probabilities across randomized input contexts. Experiments show our method outperforms standard attribution techniques in identifying hallucination causes and aligns with evidence from the model's training corpus. This work provides a unified perspective on hallucinations and a robust framework for their tracing and analysis.