# Hallucination detection from model internals

Area: [Mechanistic interpretability](https://mlsanity.com/mechinterp/index.md). HTML: https://mlsanity.com/mechinterp/hallucination-detection-from-model-internals/

## Key facts

- Hallucination detection from model internals: 23 papers on arXiv since 2023 (data through 14 Oct 2026).
- 4 of its 23 mapped papers (17%) cite “Do I Know This Entity? Knowledge Awareness and Hallucinations in Language Models” (2024, arXiv 2411.14257).
- 3 of its 23 mapped papers (13%) cite “Interpreting and Editing Vision-Language Representations to Mitigate Hallucinations” (2024, arXiv 2410.02762).
- 2 new papers 24 Sep–14 Oct.

## Most cited here

- [Do I Know This Entity? Knowledge Awareness and Hallucinations in Language Models](https://mlsanity.com/paper/2411.14257/) (2024), cited by 4 papers here
- [Interpreting and Editing Vision-Language Representations to Mitigate Hallucinations](https://mlsanity.com/paper/2410.02762/) (2024), cited by 3 papers here
- [SAFE: A Sparse Autoencoder-Based Framework for Robust Query Enrichment and Hallucination Mitigation in LLMs](https://mlsanity.com/paper/2503.03032/) (2025), cited by 2 papers here
- [Do Androids Know They're Only Dreaming of Electric Sheep?](https://mlsanity.com/paper/2312.17249/) (2023), cited by 1 papers here
- [ReDeEP: Detecting Hallucination in Retrieval-Augmented Generation via Mechanistic Interpretability](https://mlsanity.com/paper/2410.11414/) (2024), cited by 1 papers here
- [Steering LVLMs via Sparse Autoencoder for Hallucination Mitigation](https://mlsanity.com/paper/2505.16146/) (2025), cited by 1 papers here
- [SAVE: Sparse Autoencoder-Driven Visual Information Enhancement for Mitigating Object Hallucination](https://mlsanity.com/paper/2512.07730/) (2025), cited by 1 papers here
- [Toward Faithful Retrieval-Augmented Generation with Sparse Autoencoders](https://mlsanity.com/paper/2512.08892/) (2025), cited by 1 papers here
- [Dynamic Multimodal Activation Steering for Hallucination Mitigation in Large Vision-Language Models](https://mlsanity.com/paper/2602.21704/) (2026), cited by 1 papers here

## New (24 Sep–14 Oct)

- [TRACESTEER: Trace Driven Activation Steering for Hallucination Mitigation in Diffusion Language Models](https://mlsanity.com/paper/2610.18173/) (2026-10-13)
- [Where Hallucinations Live: A Cross-Architecture Circuit in VQ-Tokenized Vision-Language Models](https://mlsanity.com/paper/2609.29048/) (2026-09-24)