Wenzel Jakob: Did you know that differentiating a ...| Robocrunch AI

Did you know that differentiating a volume renderer will produce biased and noisy derivatives? Our new sampling technique fixes this, improving reconstruction of editable & relightable volumes. Joint work with @merlin_ND, Thomas Müller and Alex Keller at SIGGRAPH'22. (1/8) pic.twitter.com/gldcyEXxp7
— Wenzel Jakob (@wenzeljakob) May 13, 2022

The second ICLR'22 paper I coauthored is our Noisy Nodes work. We propose a simple noise-based regularisation that addresses oversmoothing & allows MPNNs to (finally!) reap the benefits of depth. Now with many new experiments on 2D & 3D molecular graphs. https://t.co/vcVIAwVW9D

Shared by Petar Veličković at 3/17/2022

Few-shot Learning with Noisy Labels abs: https://t.co/wGnBAoCH8D results show that TraNFS is on-par with leading FSL methods on clean support sets, yet outperforms them, by far, in the presence of label noise

Shared by AK at 4/13/2022

The Noisy Nodes method is a simple oversmoothing regularisation technique that significantly improves results for MPNNs for 3D and 2D molecular property prediction. Learn more in today’s session: https://t.co/BKGPh0mreg #ICLR2022

Shared by DeepMind at 4/26/2022

Want to know why fine-tuning can distort pretrained features (and underperform out-of-distribution)? Come to our ICLR Oral on Wednesday, 9am Pacific Time, or our poster on Tuesday at 6:30pm PT! #ICLR2022 https://t.co/yeWYhSKorS

Shared by Ananya Kumar at 4/26/2022

Thank you NeurIPS 2021 @diffprogramming workshop for giving this work the best poster award! We are excited about how AutomaticDifferentiation.jl is impacting the #julialang #sciml and #machinelearning communities. Packages are not tied to one framework. https://t.co/GsORsOc55R

Shared by Chris Rackauckas at 3/9/2022

Handling "noisy" ie. less accurate supervision is important in ML; so is using 'misaligned' supervision, eg. different granularity/labels/etc. Check out recent work here from @stevebach https://t.co/uieryYiFBk at #AISTATS2022 and Jieyu Zhang https://t.co/CtfuYZfIJv at #ICLR2022!

Shared by Alex Ratner at 4/1/2022

768