NCA-GENL Question 100
Select 4You are tasked with presenting the results of a large language model's (LLM) performance evaluation on a dataset. The evaluation includes metrics such as accuracy, inference speed, and memory utilization. Which of the following visualization techniques would be the most effective to communicate these results to stakeholders?
- A
A bar chart showing accuracy scores across different datasets
- B
A line chart depicting inference speed at varying input sizes
- C
A scatter plot comparing memory utilization to inference speed
- D
A pie chart illustrating the proportion of correct vs incorrect predictions
- E
A heatmap showing correlations between accuracy, speed, and memory utilization
Show answer and explanation
Correct answers: A, B, C, E
Explanation
Choosing the right visualization depends on the type of data and insights you want to convey. For performance evaluation metrics like accuracy, inference speed, and memory utilization, bar charts, line charts, scatter plots, and heatmaps can effectively highlight patterns, comparisons, and relationships. Pie charts, while useful for proportions, are less suitable for this type of analytical data.
- A. Correct.
Bar charts are effective for comparing discrete values such as accuracy scores across datasets, making them suitable for this scenario.
- B. Correct.
Line charts are ideal for showing trends or changes over a continuous variable, such as inference speed over varying input sizes.
- C. Correct.
Scatter plots are useful for analyzing relationships between two variables, like memory utilization and inference speed.
- D. Incorrect.
Pie charts are generally not effective for performance evaluation data, as they are more suited for showing proportions or parts of a whole, which is less relevant for this task.
- E. Correct.
Heatmaps are excellent for visualizing correlations and patterns between multiple variables, making them appropriate for presenting relationships between accuracy, speed, and memory utilization.