Michael K. Cohen

• Cohen, M. K., Kolt, N., Bengio, Y., Hadfield, G. K., & Russell, S. (2024) Regulating Advanced Artificial Agents. Science. [Brief supplementary material]

​

• Cohen, M. K., Daulton, S. & Osborne, M. A. (2022) Log-Linear-Time Gaussian Processes with Binary Tree Kernels. NeurIPS-22.

​

• Cohen, M. K., Hutter, M., & Nanda, N. (2022) Fully General Online Imitation Learning. JMLR, 23(334). (pp. 1-30)

​

• Cohen, M. K., Hutter, M., & Osborne, M. A. (2022). Advanced Artificial Agents Intervene in the Provision of Reward. AI Magazine (pp. 1-12). [Bullet-point version of paper here]

​

• Cohen, M. K., Vellambi, B., & Hutter, M. (2021) Intelligence and Unambitiousness Using Algorithmic Information Theory. IEEE Journal of Selected Areas in Information Theory. [pdf]

​

• Cohen, M. K., Catt, E., & Hutter, M. (2021) Curiosity Killed or Incapacitated the Cat and the Asymptotically Optimal Agent. IEEE Journal of Selected Areas in Information Theory.

​​

• Cohen, M. K. & Hutter, M. (2020) Pessimism About Unknown Unknowns Inspires Conservatism. In PMLR. [COLT-20]

​

• Cohen, M. K., Vellambi, B., & Hutter, M. (2020) Asymptotically Unambitious Artificial General Intelligence. In Proc. AAAI-20. [pdf]

​

• Cohen, M. K., Catt, E., & Hutter, M. (2019) A Strongly Asymptotically Optimal Agent in General Environments. In Proc. IJCAI-19. 

​

• Cohen, M. (2017). Intra-feature Random Forest Clustering. In International Workshop on Machine Learning, Optimization, and Big Data (pp. 41-49). Springer, Cham.

​

• Nemchick, D. J., Cohen, M. K., & Vaccaro, P. H. (2016). Dual hydrogen-bonding motifs in complexes formed between tropolone and formic acid. The Journal of Chemical Physics, 145(20), 204303.

​

• Nemchick, D., Cohen, M., & Vaccaro, P. (2015). Dispersion-Dominated π-Stacked Complexes Constructed on a Dynamic Scaffold. In 70th International Symposium on Molecular Spectroscopy (Vol. 1).​

​

Under review, please not distribute:

• Cohen, M., Hutter, M., Bengio, Y., & Russell, S. Don't do Anything I Wouldn't Do? KL-Regularization to an Imitative "Base Model" Does Not Implement This. [draft]

​

• Cohen, M. & Hutter, M. Imitation Learning is Probably Existentially Safe. [draft]

​

• Cohen, M. & Osborne, M. A Linear-Time, Infinite-Dimensional Extension of any Finite-Dimensional Kernel. [draft]