Understanding attribution explanations

Attribution explanations highlight specific parts of a table—such as rows, columns, or cells—that are most relevant to the answer predicted by a Table QA model. These explanations help you understand which information of the input the model considered important when predicting the answer.

Table caption: 1947 kentucky wildcats football team

Statement: The Wildcats kept the opposing team scoreless in 4 games

Game	Date	Opponent	Result	Wildcats Points	Opponents	Record
1	9999-09-20	Ole Miss	Loss	7	14	0 - 1
2	9999-09-27	Cincinnati	Win	20	0	1 - 1
4	9999-10-11	9 Georgia	Win	26	0	3 - 1 , 20
5	9999-10-18	10 Vanderbilt	Win	14	0	4 - 1 , 14
9	9999-11-15	Evansville	Win	36	0	7 - 2

In this example, the Table QA model has highlighted specific rows and cells to explain its reasoning:

• The entire rows for games 2, 4, 5, and 9 are highlighted in yellow.
• Within these rows, the "Opponents" column cells containing "0" are highlighted in green as the Statement is about "scoreless" teams.

These highlights indicate that the model identified four games where the opposing team did not score, verifying the statement as True. The yellow highlighting shows the relevant rows, while the green highlighting represents the cells containing fine-grained information needed to verify the statement.

By using different colors for highlighting, the system provides a more nuanced explanation:

• Yellow highlights (rows): Show the overall relevant context.
• Green highlights (cells): Pinpoint the exact information that directly answer the question.