https://hackerpaper.kaipfister.de/p/1706.03825/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMThttps://news.ycombinator.com/item?id=14615750urn:hackerpapers:mention:1706.03825:14615750Thu, 22 Jun 2017 22:20:01 GMT<p>I&#x27;m in no way a researcher or even an enthusiast of machine learning, but I&#x27;m pretty sure that I came across a paper posted on HN a few days ago that did exactly what you and the parent poster are describing, figuring out what pixels contributed most to some machine learning algorithm. I&#x27;ll try and see if I can find it. Edit: yep, found it. SmoothGrad: removing noise by adding noise, https://arxiv.org/abs/1706.03825 Web page with explanations and examples https://tensorflow.github.io/saliency/ I couldn&#x27;t find the HN thread, but there was no discussion as far as I remember.</p><p><a href="https://hackerpaper.kaipfister.de/p/1706.03825/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.LG