| Open problems and fundamental limitations of reinforcement learning from human feedback S Casper, X Davies, C Shi, TK Gilbert, J Scheurer, J Rando, R Freedman, ... TMLR, 2023 | 891 | 2023 |
| Rethinking machine unlearning for large language models S Liu, Y Yao, J Jia, S Casper, N Baracaldo, P Hase, Y Yao, CY Liu, X Xu, ... Nature Machine Intelligence, 1-14, 2025 | 331 | 2025 |
| Toward transparent ai: A survey on interpreting the inner structures of deep neural networks T Räuker, A Ho, S Casper, D Hadfield-Menell 2023 ieee conference on secure and trustworthy machine learning (satml), 464-483, 2023 | 311 | 2023 |
| Foundational challenges in assuring alignment and safety of large language models U Anwar, A Saparov, J Rando, D Paleka, M Turpin, P Hase, ES Lubana, ... TMLR, 2024 | 296 | 2024 |
| Scalable and transferable black-box jailbreaks for language models via persona modulation R Shah, S Pour, A Tagade, S Casper, J Rando arXiv preprint arXiv:2311.03348, 2023 | 182 | 2023 |
| Black-box access is insufficient for rigorous ai audits S Casper, C Ezell, C Siegmann, N Kolt, TL Curtis, B Bucknall, A Haupt, ... Proceedings of the 2024 ACM Conference on Fairness, Accountability, and …, 2024 | 180 | 2024 |
| Explore, establish, exploit: Red teaming language models from scratch S Casper, J Lin, J Kwon, G Culp, D Hadfield-Menell arXiv preprint arXiv:2306.09442, 2023 | 139 | 2023 |
| The ai risk repository: A comprehensive meta-review, database, and taxonomy of risks from artificial intelligence P Slattery, AK Saeri, EAC Grundy, J Graham, M Noetel, R Uuk, J Dao, ... arXiv preprint arXiv:2408.12622, 2024 | 132 | 2024 |
| International ai safety report Y Bengio, S Mindermann, D Privitera, T Besiroglu, R Bommasani, ... arXiv preprint arXiv:2501.17805, 2025 | 122 | 2025 |
| Open problems in mechanistic interpretability L Sharkey, B Chughtai, J Batson, J Lindsey, J Wu, L Bushnaq, ... TMLR, 2025 | 122 | 2025 |
| Eight methods to evaluate robust unlearning in llms A Lynch, P Guo, A Ewart, S Casper, D Hadfield-Menell arXiv preprint arXiv:2402.16835, 2024 | 122 | 2024 |
| Latent adversarial training improves robustness to persistent harmful behaviors in llms A Sheshadri, A Ewart, P Guo, A Lynch, C Wu, V Hebbar, H Sleight, ... TMLR, 2024 | 92* | 2024 |
| Open problems in technical ai governance A Reuel, B Bucknall, S Casper, T Fist, L Soder, O Aarne, L Hammond, ... TMLR, 2024 | 65 | 2024 |
| Defending against unforeseen failure modes with latent adversarial training S Casper, L Schulze, O Patel, D Hadfield-Menell TMLR, 2024 | 60 | 2024 |
| International scientific report on the safety of advanced ai (interim report) Y Bengio, S Mindermann, D Privitera, T Besiroglu, R Bommasani, ... arXiv preprint arXiv:2412.05282, 2024 | 49 | 2024 |
| Cognitive dissonance: Why do language model outputs disagree with internal representations of truthfulness? K Liu, S Casper, D Hadfield-Menell, J Andreas Proceedings of the 2023 Conference on Empirical Methods in Natural Language …, 2023 | 49 | 2023 |
| Open problems in machine unlearning for ai safety F Barez, T Fu, A Prabhu, S Casper, A Sanyal, A Bibi, A O'Gara, R Kirk, ... arXiv preprint arXiv:2501.04952, 2025 | 44 | 2025 |
| Red teaming deep neural networks with feature synthesis tools S Casper, T Bu, Y Li, J Li, K Zhang, K Hariharan, D Hadfield-Menell Advances in Neural Information Processing Systems 36, 80470-80516, 2023 | 44* | 2023 |
| Clusterability in neural networks D Filan, S Casper, S Hod, C Wild, A Critch, S Russell arXiv preprint arXiv:2103.03386, 2021 | 44 | 2021 |
| Robust feature-level adversaries are interpretability tools S Casper, M Nadeau, D Hadfield-Menell, G Kreiman Advances in Neural Information Processing Systems 35, 33093-33106, 2022 | 40 | 2022 |
Dylan Hadfield-MenellMassachusetts Institute of TechnologyVerified email at csail.mit.edu
