Thorn in a HaizeStack test for evaluating long-context adversarial robustnessgithub.com·19 pts·leonardtang·11
Degraded Polygons Raise Fundamental Questions of Neural Network Perceptionarxiv.org·1 pts·leonardtang·0
Learning the Wrong Lessons: Inserting Trojans During Knowledge Distillationarxiv.org·1 pts·leonardtang·0