Nanda, Neel

In-context Learning and Induction Heads

Olsson, C., Elhage, N., Nanda, N., Joseph, N., DasSarma, N., Henighan, T., Mann, B., Askell, A., Bai, Y., Chen, A., et al. (2022) · arXiv

Topic ModelingDomain Adaptation and Few-Shot Learning

Cited by 1,170

Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback

Bai, Y., Jones, A., Ndousse, K., Askell, A., Chen, A., DasSarma, N., Drain, D., Fort, S., Ganguli, D., Henighan, T., et al. (2022) · arXiv

Topic ModelingNatural Language Processing Techniques

Cited by 442

Predictability and Surprise in Large Generative Models

Ganguli, D., Hernandez, D., Lovitt, L., DasSarma, N., Henighan, T., Jones, A., Joseph, N., Kernion, J., Mann, B., Askell, A., et al. (2022) · arXiv

Ethics and Social Impacts of AIExplainable Artificial Intelligence (XAI)

Cited by 324

Open Problems in Mechanistic Interpretability

Sharkey, L., Chughtai, B., Batson, J., Lindsey, J., Wu, J., Bushnaq, L., Goldowsky-Dill, N., Heimersheim, S., Ortega, A., Bloom, J., et al. (2025) · arXiv

Explainable Artificial Intelligence (XAI)Neural Networks and Applications

Cited by 291

Predictability and Surprise in Large Generative Models

Ganguli, D., Hernandez, D., Lovitt, L., Askell, A., Bai, Y., Chen, A., Conerly, T., Dassarma, N., Drain, D., Elhage, N., et al. (2022) · 2022 ACM Conference on Fairness Accountability and Transparency

Ethics and Social Impacts of AIExplainable Artificial Intelligence (XAI)

Cited by 207

AtP*: An efficient and scalable method for localizing LLM behaviour to components

Kramár, J., Lieberum, T., Shah, R., & Nanda, N. (2024) · arXiv

Explainable Artificial Intelligence (XAI)Software Testing and Debugging Techniques

Cited by 120

Refusal in Language Models Is Mediated by a Single Direction

Arditi, A., Obeso, O., Syed, A., Paleka, D., Panickssery, N., Gurnee, W., & Nanda, N. (2024) · Advances in Neural Information Processing Systems 37

Language, Discourse, Communication StrategiesLinguistics and Language Analysis

Cited by 70

Emergent Linear Representations in World Models of Self-Supervised Sequence Models

Nanda, N., Lee, A., & Wattenberg, M. (2023) · Proceedings of the 6th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP

Explainable Artificial Intelligence (XAI)Artificial Intelligence in Games

Cited by 55

Emergent Linear Representations in World Models of Self-Supervised Sequence Models

Nanda, N., Lee, A., & Wattenberg, M. (2023) · arXiv

Explainable Artificial Intelligence (XAI)Artificial Intelligence in Games

Cited by 55

Progress measures for grokking via mechanistic interpretability

Nanda, N., Chan, L., Lieberum, T., Smith, J., & Steinhardt, J. (2023) · arXiv

Explainable Artificial Intelligence (XAI)Formal Methods in Verification

Cited by 51

Thought Anchors: Which LLM Reasoning Steps Matter?

Bogdan, P. C., Macar, U., Nanda, N., & Conmy, A. (2025) · arXiv

Explainable Artificial Intelligence (XAI)Topic Modeling

Cited by 24

Finding Neurons in a Haystack: Case Studies with Sparse Probing

Gurnee, W., Nanda, N., Pauly, M., Harvey, K., Troitskii, D., & Bertsimas, D. (2023) · arXiv

Explainable Artificial Intelligence (XAI)Topic Modeling

Cited by 15

Refusal in Language Models Is Mediated by a Single Direction

Arditi, A., Obeso, O., Syed, A., Paleka, D., Panickssery, N., Gurnee, W., & Nanda, N. (2024) · arXiv

Adversarial Robustness in Machine LearningSecurity and Verification in Computing

Cited by 9

Chain-of-Thought Reasoning In The Wild Is Not Always Faithful

Arcuschin, I., Janiak, J., Krzyzanowski, R., Rajamanoharan, S., Nanda, N., & Conmy, A. (2025) · arXiv

Explainable Artificial Intelligence (XAI)Topic Modeling

Cited by 6

Improving Dictionary Learning with Gated Sparse Autoencoders

Rajamanoharan, S., Conmy, A., Smith, L., Lieberum, T., Varma, V., Kramár, J., Shah, R., & Nanda, N. (2024) · arXiv

Explainable Artificial Intelligence (XAI)Topic Modeling

Cited by 3

Confidence Regulation Neurons in Language Models

Stolfo, A., Wu, B., Gurnee, W., Belinkov, Y., Song, X., Sachan, M., & Nanda, N. (2024) · Advances in Neural Information Processing Systems 37

Neurobiology of Language and BilingualismTopic Modeling

Cited by 2

Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2

Lieberum, T., Rajamanoharan, S., Conmy, A., Smith, L., Sonnerat, N., Varma, V., Kramár, J., Dragan, A., Shah, R., & Nanda, N. (2024) · arXiv

Explainable Artificial Intelligence (XAI)Topic Modeling

Cited by 2

Because we have LLMs, we Can and Should Pursue Agentic Interpretability

Kim, B., Hewitt, J., Nanda, N., Fiedel, N., & Tafjord, O. (2025) · arXiv

Explainable Artificial Intelligence (XAI)Ethics and Social Impacts of AI

Cited by 1

Convergent Linear Representations of Emergent Misalignment

Soligo, A., Turner, E., Rajamanoharan, S., & Nanda, N. (2025) · arXiv

Domain Adaptation and Few-Shot LearningTopic Modeling

Cited by 0

Model Organisms for Emergent Misalignment

Turner, E., Soligo, A., Taylor, M., Rajamanoharan, S., & Nanda, N. (2025) · arXiv

Ethics and Social Impacts of AIAdversarial Robustness in Machine Learning

Cited by 0