RL fine-tuning with a counterfactual mention/influence reward raises LLM self-explanation faithfulness (Phi-CCT) from near zero to ~0.66 in-distribution for two 8B models, with partial transfer to held-out tasks.
Assessing Gender Bias in LLMs: Comparing LLM Outputs with Human Perceptions and Official Statistics
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
This study investigates gender bias in large language models (LLMs) by comparing their gender perception to that of human respondents, U.S. Bureau of Labor Statistics data, and a 50% no-bias benchmark. We created a new evaluation set using occupational data and role-specific sentences. Unlike common benchmarks included in LLM training data, our set is newly developed, preventing data leakage and test set contamination. Five LLMs were tested to predict the gender for each role using single-word answers. We used Kullback-Leibler (KL) divergence to compare model outputs with human perceptions, statistical data, and the 50% neutrality benchmark. All LLMs showed significant deviation from gender neutrality and aligned more with statistical data, still reflecting inherent biases.
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Training Large Language Models for Self-Explanation Faithfulness
RL fine-tuning with a counterfactual mention/influence reward raises LLM self-explanation faithfulness (Phi-CCT) from near zero to ~0.66 in-distribution for two 8B models, with partial transfer to held-out tasks.