Self-Blinding and Counterfactual Self-Simulation Mitigate Biases and Sycophancy in Large Language Models

Published in arXiv, 2026

Recommended citation: Christian, B. & Mazor, M. (2026). Self-Blinding and Counterfactual Self-Simulation Mitigate Biases and Sycophancy in Large Language Models. /files/papers/christian2026blinding.pdf