Update README.md

README.md

@@ -15,7 +15,10 @@ permutations and standard Young tableaux.
 
 The goal of this dataset is to see whether a model can learn the RSK algorithm. 
 That is, for a fixed \\(n\\) the model is provided with a pair of standard Young tableaux 
-\\(\lambda_1, \lambda_2\\) of size \\(n\\) and required to predict the corresponding permutation. 
+\\(\lambda_1, \lambda_2\\) of size \\(n\\) and required to predict the corresponding permutation
+(note that the RSK algorithm is usually introduced as a map from permutations to pairs of
+standard Young tableaux but we have found that it is easier to design neural 
+architectures that output permutations and have hence reversed the bijection for this dataset).
 Although the algorithm is known, it would be significant for a model to learn this 
 correspondence due to the the intricate combinatorial rules involved. Notably, 
 the RSK correspondence can be used to find the length of the longest increasing subsequence of a permutation,