Professions (Bolukbasi et al.)

anchor

Findings

• IC-1078LLaMA models (7B through 65B) exhibit gender bias in language generation, coreference resolution, and sentence likelihood, with stereotypical associations driving predictions [eval]
• IC-1079Causal tracing reveals that mid-upper MLP layers (layers 18–25 in 7B) are the primary mediators of stereotypical gender bias in LLaMA, while the last layers show negative coefficients that counter the bias [eval]