Figure 2
A histogram showing the distribution of average sentence length by label. The x-axis represents the average sentence length, ranging from 0 to 175. The y-axis represents the density, ranging from 0 to 0.08. The histogram has two sets of bars, one in blue labeled as AI and one in orange labeled as Non-AI. The blue bars for AI are concentrated around lower average sentence lengths, with a peak density around 20. The orange bars for Non-AI are more spread out, with a peak density around 25. Both distributions show a right-skewed pattern, with most values clustered at lower sentence lengths and tapering off as the sentence length increases. The blue bars for AI are generally taller and more concentrated than the orange bars for Non-AI, indicating that AI-generated sentences tend to be shorter on average. All values are approximated.Distribution of avg_sentence_len by label