9/ The Unreasonable Effectiveness of Easy Training Data for Hard Tasks – suggests that language models often generalize well from easy to hard data, i.e., easy-to-hard generalization.
Easy Training Data Enables Hard Task Generalization
By
–
By
–
9/ The Unreasonable Effectiveness of Easy Training Data for Hard Tasks – suggests that language models often generalize well from easy to hard data, i.e., easy-to-hard generalization.