We have 2 of 5 papers
We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.
Not the right person? Other researchers publish under this name.
A truth probe fitted where truthful reporting and a task's prescribed action coincide cannot distinguish those targets from its fitting labels alone, and this failure of semantic identification perfect aliasing is called failure of semantic identification perfect aliasing.
The mechanism is a halt vector: a difference-of-means direction at layer 18 of this model whose steering strength controls how long it thinks, while a replicated value axis does nothing, and what works is reconstructing the whole steered activation with those dimensions pinned to their natural values.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
