Rejection ≠ Failure: A NeurIPS-Rejected Paper Wins the AISTATS 2025 Test of Time Award
-
A paper that received review scores of 8/8/7 from NeurIPS over a decade ago—but was rejected and later published at AISTATS — had been honored with the AISTATS 2025 Test of Time Award.
It has since been cited over 3,000 times according to Google Scholar—an impressive testament to its lasting impact.
The author, Professor Saining Xie, shared that this was the first paper he submitted as a PhD student. The sting of rejection stayed with him for years, but this award has brought closure. In his own words, he hopes this story reminds students and early-career researchers that persistence matters, even in the face of disheartening reviews.“I wouldn’t call conferences a lottery, but a bit of perseverance does go a long way.” – Saining Xie
If you’ve had a paper that was initially rejected but later turned out to be influential—or if you’ve reviewed work that only gained recognition years later—we’d love to hear your story and reflections.
Reference:
Lee, Chen-Yu, Saining Xie, Patrick Gallagher, Zhengyou Zhang, and Zhuowen Tu.
“Deeply-supervised nets.” In Artificial Intelligence and Statistics, pp. 562–570. PMLR, 2015.The abstract of the paper:
-
Well recognized! FINALLY!
