Search Constraints

Start Over
Filtering by: Creator Taylor, Dennis William Remove constraint Creator: Taylor, Dennis William Committee Member Li, Fuxin Remove constraint Committee Member: Li, Fuxin License Attribution 4.0 (CC BY 4.0) Remove constraint License: Attribution 4.0 (CC BY 4.0)

Search Results

  • A Comparative Study of Non-Normal Distributions in Continuous Dropout

    Creator:
    Guyer, Alexander D.
    Abstract:
    Recent studies have shown that novel continuous dropout methods can be viewed as a Bayesian interpretation of model parameters, though most such studies have shown results using normal distributions. As the posterior distributions over neural network nodes and parameters are intractable, given that they are a result of artificial construction...
    Resource Type:
    Honors College Thesis
    Full Text:
    : William Smart Recent studies have shown that novel continuous dropout methods can be viewed as a Bayesian

  • On Avoiding Power-Seeking by Artificial Intelligence

    Creator:
    Turner, Alexander M.
    Abstract:
    We do not know how to align a very intelligent AI agent's behavior with human interests. I investigate whether—absent a full solution to this AI alignment problem—we can build smart {\ai} agents which have limited impact on the world, and which do not autonomously seek power. In this thesis, I...
    Resource Type:
    Dissertation