Asked about Reading papers without reading papers
What makes you sceptical of a paper's result?
An answer that works
A weak baseline first, since most disputes are really about baselines. Then single-seed results with no variance reported, small effects framed as large, and missing ablations, which often mean a component did not help.
Then they push further
Give me a specific phrase you would stop on.
'We use the default settings for the baseline', which usually means the method got a hyperparameter sweep the baseline did not. Also 'results are averaged over multiple runs' without saying how many, and an improvement reported only as a percentage, which hides whether the absolute change is half a point.
What gets marked up
A strong answer names the sentence in the paper that triggers the check, not the abstract category. Naming baselines and variance is passable; saying what you would look up next is what separates it.
Reading an answer is not being able to give one. The section that teaches this is Reading papers without reading papers, 5 hours of Reading papers.
Learn this properlyAlso asked about Reading papers