How to extract Correlation of Coefficients table from models like glm?

After running a statistical model in R (e.g., glm, lm, lme4::lmer, etc.), I run the summary() command with corr=TRUE to get the Correlation of Coefficients table. It features a matrix of correlations for the intercept and the independent variables in the given model. What I would like to do is extract the columns that make those correlations possible. For instance, if the model is x ~ a + b, and the table looks like this...

 (Intercept)   a   a    0.##      b    0.##     0.##

I want to extract the columns that allow for the correlation between a and b.

I have found various commands to extract fitted values and residuals, but nothing to yet that would give me values for a and b that correspond to their correlations in that table.