Manipulating Feature Visualizations with Gradient Slingshots
@dilya.bsky.social Marina MC HΓΆhne, Alexander Warnecke @lpirch.bsky.social Klaus-Robert MΓΌller @rieck.mlsec.org @slapuschkin.bsky.social @kirillbykov.bsky.social
π
Manipulating Feature Visualizations with Gradient Slingshots
@dilya.bsky.social Marina MC HΓΆhne, Alexander Warnecke @lpirch.bsky.social Klaus-Robert MΓΌller @rieck.mlsec.org @slapuschkin.bsky.social @kirillbykov.bsky.social
π
Predicted temperature at the peak of the European Heat Wave 2006
AI predicts rain. We predict trouble!
Today, Erik presents a novel attack on Google's latest AI weather model at #CCS2025. By changing only 0.1% of the observations, the attack can fabricate or suppress the prediction of extreme events, from hurricanes π to heat waves π₯
1/4 @bifold.berlin
Photogenic as always π
Members of the MLSEC team (mlsec.org)
Great to be at @satml.org with several members of my team from @bifold.berlin and @tuberlin.bsky.social. We are having a blast with exciting discussions and talks on trustworthy AI! #SaTML25
riding to SaTML25 in Copenhagen
initial commit. add README and a pic of us driving to #SaTML25 this morning π²βοΈ