Zobrazeno 1 - 10
of 403
pro vyhledávání: '"Sklar, Michael"'
Autor:
Thompson, T. Ben, Sklar, Michael
Many publicly available language models have been safety tuned to reduce the likelihood of toxic or liability-inducing text. Users or security analysts attempt to jailbreak or redteam these models with adversarial prompts which cause compliance with
Externí odkaz:
http://arxiv.org/abs/2407.17447
Feature visualization, also known as "dreaming", offers insights into vision models by optimizing the inputs to maximize a neuron's activation or other internal component. However, dreaming has not been successfully applied to language models because
Externí odkaz:
http://arxiv.org/abs/2402.01702
Contemporary scientific research is a distributed, collaborative endeavor, carried out by teams of researchers, regulatory institutions, funding agencies, commercial partners, and scientific bodies, all interacting with each other and facing differen
Externí odkaz:
http://arxiv.org/abs/2307.03748
We present a proof-by-simulation framework for rigorously controlling a trial design's operating characteristics over continuous regions of parameter space. We show that proof-by-simulation can achieve two major goals: (1) Calibrate a design for prov
Externí odkaz:
http://arxiv.org/abs/2212.10042
Autor:
Sklar, Michael
This PhD thesis covers breakthroughs in several areas of adaptive experiment design: (i) (Chapter 2) Novel clinical trial designs and statistical methods in the era of precision medicine. (ii) (Chapter 3) Multi-armed bandit theory, with applications
Externí odkaz:
http://arxiv.org/abs/2205.09369
Consider the relationship between a regulator (the principal) and an experimenter (the agent) such as a pharmaceutical company. The pharmaceutical company wishes to sell a drug for profit, whereas the regulator wishes to allow only efficacious drugs
Externí odkaz:
http://arxiv.org/abs/2205.06812
Autor:
Abbott, Megan, Pereira, Sergio M., Sanders, Noah, Girard, Martin, Sankar, Ashwin, Sklar, Michael C.
Publikováno v:
In British Journal of Anaesthesia August 2024 133(2):424-436
Autor:
Quickfall, Danica, Sklar, Michael C., Tomlinson, George, Orchanian-Cheff, Ani, Goligher, Ewan C.
Publikováno v:
In eClinicalMedicine February 2024 68
We consider the problem of evaluating designs for a two-arm randomized experiment with the criterion being the power of the randomization test for the one-sided null hypothesis. Our evaluation assumes a response that is linear in one observed covaria
Externí odkaz:
http://arxiv.org/abs/2008.05980
We present an optimized rerandomization design procedure for a non-sequential treatment-control experiment. Randomized experiments are the gold standard for finding causal effects in nature. But sometimes random assignments result in unequal partitio
Externí odkaz:
http://arxiv.org/abs/1905.03337