Results 41 to 50 of about 7,745 (121)

When Choosing Plausible Alternatives, Clever Hans can be Clever [PDF]

open access: yesarXiv, 2019
Pretrained language models, such as BERT and RoBERTa, have shown large improvements in the commonsense reasoning benchmark COPA. However, recent work found that many improvements in benchmarks of natural language understanding are not due to models learning the task, but due to their increasing ability to exploit superficial cues, such as tokens that ...
arxiv  

UWOMJ Volume 10, No 4 1940 [PDF]

open access: yes, 1940
Schulich School of Medicine & Dentistryhttps://ir.lib.uwo.ca/uwomj/1116/thumbnail ...
Western University
core   +1 more source

Lattice Fluctuations Spectrum Calculation for P-dihlorbenzol Nanocrystals [PDF]

open access: yesarXiv, 2011
We have measured spectra of lattice fluctuations in p-dihlorbenzol nanoparticles of ~300nm size. Calculations of lattice frequencies and of these fluctuations in nanoparticles are done. It is shown that with reduction of nanoparticles sizes the spectrum of superficial fluctuations becomes prevailing.
arxiv  

Traumatic funicular phlebitis of the thoracic wall resembling Mondor's disease: a case report [PDF]

open access: yes, 2011
Introduction Mondor's disease is a peculiar form of thrombophlebitis, involving a superficial vein in the subcutaneous fat of the breast or anterior chest wall. Case presentation The author presents a case of a 35-year-old male Japanese patient with cord-
A Ichinose   +8 more
core   +3 more sources

Learning Robust Representations by Projecting Superficial Statistics Out [PDF]

open access: yesarXiv, 2019
Despite impressive performance as evaluated on i.i.d. holdout data, deep neural networks depend heavily on superficial statistics of the training data and are liable to break under distribution shift. For example, subtle changes to the background or texture of an image can break a seemingly powerful classifier.
arxiv  

Extracting and Understanding the Superficial Knowledge in Alignment [PDF]

open access: yesarXiv
Alignment of large language models (LLMs) with human values and preferences, often achieved through fine-tuning based on human feedback, is essential for ensuring safe and responsible AI behaviors. However, the process typically requires substantial data and computation resources. Recent studies have revealed that alignment might be attainable at lower
arxiv  

The role of a malignancy in the treatment of a venous thromboembolism and superficial thrombophlebitis [PDF]

open access: yes, 2017
Cancer patients have a manifold risk of suffering from both thrombotic events and anticoagulation-related bleeding complications. For this reason, knowledge of their adequate medication is crucial.
Taanila, Anette
core  

Home - About - Disclaimer - Privacy