The paper presents a study on the degrees of freedom of the lasso in the framework of Stein's unbiased risk estimation (SURE). It shows that the number of nonzero coefficients in the lasso solution is an unbiased estimate of the degrees of freedom, which is asymptotically consistent. This result allows for the construction of model selection criteria such as $ C_p $, AIC, and BIC, which can be used to select the optimal lasso fit efficiently. The degrees of freedom for the lasso is defined as the expected number of nonzero predictors in the model. The paper also discusses the importance of the degrees of freedom in model assessment and selection, and shows that the unbiased estimator of the degrees of freedom can be used to construct these criteria. The study further demonstrates that the unbiased estimator is consistent and provides a principled approach to model selection. The paper also addresses the issue of model selection in the presence of high-dimensional data and shows that the lasso can be used to select the optimal model with computational efficiency. The results are illustrated with numerical experiments and simulations, and the paper concludes with a discussion on the implications of the findings for model selection and the use of the lasso in statistical modeling.The paper presents a study on the degrees of freedom of the lasso in the framework of Stein's unbiased risk estimation (SURE). It shows that the number of nonzero coefficients in the lasso solution is an unbiased estimate of the degrees of freedom, which is asymptotically consistent. This result allows for the construction of model selection criteria such as $ C_p $, AIC, and BIC, which can be used to select the optimal lasso fit efficiently. The degrees of freedom for the lasso is defined as the expected number of nonzero predictors in the model. The paper also discusses the importance of the degrees of freedom in model assessment and selection, and shows that the unbiased estimator of the degrees of freedom can be used to construct these criteria. The study further demonstrates that the unbiased estimator is consistent and provides a principled approach to model selection. The paper also addresses the issue of model selection in the presence of high-dimensional data and shows that the lasso can be used to select the optimal model with computational efficiency. The results are illustrated with numerical experiments and simulations, and the paper concludes with a discussion on the implications of the findings for model selection and the use of the lasso in statistical modeling.