2 papers
cs.AI2026
One Probe Won't Catch Them All: Towards Targeted Deception Detection
Vikram Natarajan, Devina Jain, Shivam Arora +2
Linear probes are a promising approach for monitoring AI systems for deceptive behaviour. Previous work has shown that a linear classifier trained on a contrastive instruction pair…
cs.LG2026
Sustained Gradient Alignment Mediates Subliminal Learning in a Multi-Step Setting: Evidence from MNIST Auxiliary Logit Distillation Experiment
Chayanon Kitkana, Shivam Arora
In the MNIST auxiliary logit distillation experiment, a student can acquire an unintended teacher trait despite distilling only on no-class logits through a phenomenon called subli…