Brown University

About: Brown University is a education organization based out in Providence, Rhode Island, United States. It is known for research contribution in the topics: Population & Poison control. The organization has 35778 authors who have published 90896 publications receiving 4471489 citations. The organization is also known as: brown.edu & Brown.

Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference

Abstract: A machine learning system can score well on a given test set by relying on heuristics that are effective for frequent example types but break down in more challenging cases. We study this issue within natural language inference (NLI), the task of determining whether one sentence entails another. We hypothesize that statistical NLI models may adopt three fallible syntactic heuristics: the lexical overlap heuristic, the subsequence heuristic, and the constituent heuristic. To determine whether models have adopted these heuristics, we introduce a controlled evaluation set called HANS (Heuristic Analysis for NLI Systems), which contains many examples where the heuristics fail. We find that models trained on MNLI, including BERT, a state-of-the-art model, perform very poorly on HANS, suggesting that they have indeed adopted these heuristics. We conclude that there is substantial room for improvement in NLI systems, and that the HANS dataset can motivate and measure progress in this area

STARD 2015: An Updated List of Essential Items for Reporting Diagnostic Accuracy Studies.

Abstract: Incomplete reporting has been identified as a major source of avoidable waste in biomedical research. Essential information is often not provided in study reports, impeding the identification, critical appraisal, and replication of studies. To improve the quality of reporting of diagnostic accuracy studies, the Standards for Reporting of Diagnostic Accuracy Studies (STARD) statement was developed. Here we present STARD 2015, an updated list of 30 essential items that should be included in every report of a diagnostic accuracy study. This update incorporates recent evidence about sources of bias and variability in diagnostic accuracy and is intended to facilitate the use of STARD. As such, STARD 2015 may help to improve completeness and transparency in reporting of diagnostic accuracy studies.

Renewable neural implant device and method

Abstract: Refillable immunoisolatory neurological therapy devices for local and controlled delivery of a biologically active factor to the brain of a patient. The devices include a cell chamber adapted for infusion with nsecretory cells and having at least one semipermeable or permselective surface across which biologically active factors secreted by the cells can be delivered to the brain. The devices also include means for introducing secretory cells into the cell chamber, and means for renewing the cells or cell medium.

Accuracy and stability of integration algorithms for elastoplastic constitutive relations

Abstract: An analysis of accuracy and stability of algorithms for the integration of elastoplastic constitutive relations is carried out in this paper. Reference is made to a very general internal variable formulation of plasticity and to two families of algorithms that generalize the well-known trapezoidal and midpoint rules to fit the present context. Other integration schemes such as the radial return, mean normal and closest point procedures are particular cases of this general formulation. The meaning of first and second-order accuracy in the presence of the plastic consistency condition is examined in detail, and the criteria derived are used to identify two second-order accurate members of the proposed algorithms. A general methodology is also derived whereby the numerical stability properties of integration schemes can be systematically assessed. With the aid of this methodology, the generalized midpoint rule is seen to have far better stability properties than the generalized trapezoidal rule. Finally, numerical examples are presented that illustrate the performance of the algorithms.