Stephan Rabanser
@steverab
PhD candidate @UofT and @VectorInst. Soon: PostDoc @Princeton. Reliable, safe, trustworthy machine learning.
ID: 138821636
https://rabanser.dev 30-04-2010 18:05:23
10,10K Tweet
447 Takipçi
338 Takip Edilen
Trustworthy ML Initiative (TrustML) ICML Conference We have some cool work on manipulating model uncertainty / model abstention mechanisms and reliably detecting when such tampering occurs via zero knowledge proofs! x.com/steverab/statu…