Stepwiselm gives error "Maximum variable size allowed by the program is exceeded". Can I override this?

I am trying a stepwiselm regression and get the error "Maximum variable size allowed by the program is exceeded". Trying poly3 terms, but I have 48 predictors (x values).

I have the x values loaded into x_values, an data type double (6000 x 48). The y values are similarly loaded into y_values (6000 x 1). I want to specify that the upper bound limitation on the model can be cubed.

I have successfully executed the quadratic version: stepwiselm(x_values, y_values, 'upper', 'quadratic'). But when I try it with poly: stepwiselm(x_values, y_values, 'upper', 'poly333..(48 3's go here)... 333'), I get the error that maximum variable size allowed is exceeded. Is there a way to disable, or get around this?

The full error is below: ------ Error using repmat Maximum variable size allowed by the program is exceeded.

Error in ndgrid (line 73) varargout{i} = repmat(x,s);

Error in classreg.regr.modelutils.model2terms (line 85) [powers{1:nvars}] = ndgrid(powers{:});

Error in classreg.regr.FormulaProcessor (line 361) f.terms = classreg.regr.modelutils.model2terms(alias,predVars,hasIntercept);

Error in classreg.regr.LinearFormula (line 48) f = [e-mail](varargin{:});

Error in LinearModel.stepwise (line 90) upper = classreg.regr.LinearFormula(upper,start.VariableNames,start.ResponseName,start.HasIntercept,start.Link);

Error in stepwiselm (line 100) model = LinearModel.stepwise(X,varargin{:}); ------ Thank you for your help!

Jeff