Funny rephrasings of LLMs in academic papers
While reading and reviewing academic papers I have found hilarious sentences that, most likely, were originally written differently, and were rephrased by a LLM like ChatGPT. I wanted to share this to remind that one must carefully review the text these LLMs produced before submitting a paper.
Deep Learning / Machine Learning
"Field of sensation" -> Receptive field
"Self-taught learning" -> Self-supervised learning
"Input characteristics" -> Input features
"Machine wisdom" -> Machine learning
"Device learning" -> Machine learning
"Image dispensation" -> Image processing
"Model structure" -> Model architecture
"Gradient disappearance" -> Vanishing gradients
"Gradient explosive" -> Exploding gradients
"Damage function" -> Loss function
"Prejudiced cross entropy" -> Weighted cross entropy
"Deceased loss function" -> Dice (mistyped as "die") loss function
"Die harm" -> Dice loss
MRI
"Data gathering" -> Data acquisition
"Mental pictures" -> Brain images