NeurIPS 2021 Reveals Outstanding Paper and Test of Time Award Winners
NeurIPS 2021 officially announces the recipients of its Outstanding Paper Awards, Test of Time Award, and the inaugural Datasets and Benchmarks Track Best Paper Awards. Six papers earn top honors this year for their exceptional clarity, creativity, and potential for lasting impact.
NeurIPS 2021 officially announces the recipients of its prestigious awards just days before the conference begins. The organization honors six papers with Outstanding Paper Awards, recognizing their exceptional clarity, insight, creativity, and potential for lasting impact on the field. A separate Test of Time Award highlights historically significant research, while the new Datasets and Benchmarks Track introduces its first-ever Best Paper Awards.
Dedicated award committees carefully evaluate the submissions to select this year's winners. The Outstanding Paper Award committee includes prominent researchers such as Alice Oh, Daniel Hsu, Emma Brunskill, Kilian Weinberger, and Yisong Yue. Meanwhile, the Test of Time Award committee features Joelle Pineau, Léon Bottou, Max Welling, and Ulrike von Luxburg, with additional support from Nati Srebro and various subject-matter experts.
Among the highlighted Outstanding Paper Award recipients is "A Universal Law of Robustness via Isoperimetry" by Sébastien Bubeck and Mark Sellke. This specific paper offers a theoretical model explaining why modern deep networks require significantly more parameters than expected to smoothly fit training data. The research demonstrates that parameter requirements scale with both the number of training examples and data dimensionality, providing an elegant theory that aligns with empirical observations about robust model sizes.