| Ethical and social risks of harm from language models L Weidinger, J Mellor, M Rauh, C Griffin, J Uesato, PS Huang, M Cheng, ... arXiv preprint arXiv:2112.04359, 2021 | 998 | 2021 |
| Taxonomy of risks posed by language models L Weidinger, J Uesato, M Rauh, C Griffin, PS Huang, J Mellor, A Glaese, ... Proceedings of the 2022 ACM Conference on Fairness, Accountability, and …, 2022 | 612 | 2022 |
| Three factors influencing minima in sgd S Jastrzębski, Z Kenton, D Arpit, N Ballas, A Fischer, Y Bengio, A Storkey arXiv preprint arXiv:1711.04623, 2017 | 532 | 2017 |
| Alignment of language agents Z Kenton, T Everitt, L Weidinger, I Gabriel, V Mikulik, G Irving arXiv preprint arXiv:2103.14659, 2021 | 164 | 2021 |
| A systematic comparison of bayesian deep learning robustness in diabetic retinopathy tasks A Filos, S Farquhar, AN Gomez, TGJ Rudner, Z Kenton, L Smith, ... arXiv preprint arXiv:1912.10481, 2019 | 130 | 2019 |
| On the relation between the sharpest directions of DNN loss and the SGD step length S Jastrzębski, Z Kenton, N Ballas, A Fischer, Y Bengio, A Storkey arXiv preprint arXiv:1807.05031, 2018 | 129 | 2018 |
| Specification gaming: the flip side of AI ingenuity V Krakovna, J Uesato, V Mikulik, M Rahtz, T Everitt, R Kumar, Z Kenton, ... DeepMind Blog 3, 2020 | 113 | 2020 |
| Ethical and social risks of harm from language models. arXiv L Weidinger, J Mellor, M Rauh, C Griffin, J Uesato, PS Huang, M Cheng, ... arXiv preprint arXiv:2112.04359 10, 2021 | 76 | 2021 |
| Imitating interactive intelligence J Abramson, A Ahuja, I Barr, A Brussee, F Carnevale, M Cassin, ... arXiv preprint arXiv:2012.05672, 2020 | 74 | 2020 |
| Goal misgeneralization: Why correct specifications aren't enough for correct goals R Shah, V Varma, R Kumar, M Phuong, V Krakovna, J Uesato, Z Kenton arXiv preprint arXiv:2210.01790, 2022 | 66 | 2022 |
| The ethics of advanced ai assistants I Gabriel, A Manzini, G Keeling, LA Hendricks, V Rieser, H Iqbal, ... arXiv preprint arXiv:2404.16244, 2024 | 50 | 2024 |
| Explaining grokking through circuit efficiency V Varma, R Shah, Z Kenton, J Kramár, R Kumar arXiv preprint arXiv:2309.02390, 2023 | 44 | 2023 |
| Finding flatter minima with sgd S Jastrzębski, Z Kenton, D Arpit, N Ballas, A Fischer, Y Bengio, A Storkey | 40 | 2018 |
| The squeezed limit of the bispectrum in multi-field inflation Z Kenton, DJ Mulryne Journal of Cosmology and Astroparticle Physics 2015 (10), 018, 2015 | 39 | 2015 |
| D-brane potentials in the warped resolved conifold and natural inflation Z Kenton, S Thomas Journal of High Energy Physics 2015 (2), 1-42, 2015 | 39 | 2015 |
| Discovering agents Z Kenton, R Kumar, S Farquhar, J Richens, M MacDermott, T Everitt Artificial Intelligence 322, 103963, 2023 | 34 | 2023 |
| Width of minima reached by stochastic gradient descent is influenced by learning rate to batch size ratio S Jastrzębski, Z Kenton, D Arpit, N Ballas, A Fischer, Y Bengio, A Storkey Artificial Neural Networks and Machine Learning–ICANN 2018: 27th …, 2018 | 30 | 2018 |
| Generalizing from a few environments in safety-critical reinforcement learning Z Kenton, A Filos, Y Gal, O Evans Safe Machine Learning workshop at ICLR, 2019 | 26* | 2019 |
| Benchmarking Bayesian deep learning with diabetic retinopathy diagnosis A Filos, S Farquhar, AN Gomez, TGJ Rudner, Z Kenton, L Smith, ... Preprint at https://arxiv. org/abs/1912.10481, 2019 | 24 | 2019 |
| Ethical and social risks of harm from Language Models. arXiv 2021 L Weidinger, J Mellor, M Rauh, C Griffin, J Uesato, PS Huang, M Cheng, ... arXiv preprint arXiv:2112.04359, 2021 | 23 | 2021 |
Amos StorkeyProfessor of Machine Learning and AI, School of Informatics, University of Edinburgh, UKAdresse e-mail validée de ed.ac.uk