MATS Fellow:
Niels uit de Bos
Authors:
Niels uit de Bos, Adrià Garriga-Alonso
Citations
Abstract:
Circuits are supposed to accurately describe how a neural network performs a specific task, but do they really? We evaluate three circuits found in the literature (IOI, greater-than, and docstring) in an adversarial manner, considering inputs where the circuit's behavior maximally diverges from the full model. Concretely, we measure the KL divergence between the full model's output and the circuit's output, calculated through resample ablation, and we analyze the worst-performing inputs. Our results show that the circuits for the IOI and docstring tasks fail to behave similarly to the full model even on completely benign inputs from the original task, indicating that more robust circuits are needed for safety-critical applications.
What Happens When Superhuman AIs Compete for Control?
Authors:
Steven Veld
Date:
January 11, 2026
Citations:
0
AI Futures Model: Timelines & Takeoff
Authors:
Brendan Halstead, Alex Kastner
Date:
December 30, 2025
Citations:
0
The MATS Program is an independent research and educational initiative connecting emerging researchers with mentors in AI alignment, governance, and security.
Each MATS cohort runs for 12 weeks in Berkeley, California, followed by an optional 6–12 month extension in London for selected scholars.