Goto

Collaborating Authors

 Alhambra


Mechanistic Interpretability Needs Philosophy

Williams, Iwan, Oldenburg, Ninell, Dhar, Ruchira, Hatherley, Joshua, Fierro, Constanza, Rajcic, Nina, Schiller, Sandrine R., Stamatiou, Filippos, Søgaard, Anders

arXiv.org Artificial Intelligence

Mechanistic interpretability (MI) aims to explain how neural networks work by uncovering their underlying causal mechanisms. As the field grows in influence, it is increasingly important to examine not just models themselves, but the assumptions, concepts and explanatory strategies implicit in MI research. We argue that mechanistic interpretability needs philosophy: not as an afterthought, but as an ongoing partner in clarifying its concepts, refining its methods, and assessing the epistemic and ethical stakes of interpreting AI systems. Taking three open problems from the MI literature as examples, this position paper illustrates the value philosophy can add to MI research, and outlines a path toward deeper interdisciplinary dialogue.


New Jersey Man Shot Dead After Torture Over PlayStation, Suspect Arrested

International Business Times

Authorities in New Jersey arrested a man accused of kidnapping and murdering a 20-year-old trying to buy a PlayStation video game console advertised online, reports said Saturday. Rufus Thompson, 29, was arrested Saturday morning in Trenton and was charged with murder, felony murder, robbery, kidnapping and weapons offenses in the death of Danny Diaz-Delgado. The victim's body was discovered March 24 near the banks of Assunpink Creek in Hamilton Township, the Trentonian reported. Investigators said Diaz-Delgado had been shot at least seven times, with wounds in his head, torso and one leg. His hands were tied behind his back and tape was wrapped around his face.