Self-rewarding correction for mathematical reasoning

Open in new window