Miles Redcliffe, and Matthew Blackwood. 2026. “Improving Interpretability in Large Language Models through Layer Wise Relevance Propagation for Identifying Latent Reasoning Biases”. International Journal of Artificial Intelligence Research 1 (2). https://doi.org/10.66280/ijair.v1i2.142.