Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large Language Models. Lindsey, J., van der Weij, T., Nanda, N., & others arXiv preprint arXiv:2601.14004, 2026.
bibtex   
@article{lindsey2026actionableMI,
  title={Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large Language Models},
  author={Lindsey, Jack and van der Weij, Tom and Nanda, Neel and others},
  journal={arXiv preprint arXiv:2601.14004},
  year={2026}
}

Downloads: 0