Step Back to Leap Forward: Self-Backtracking for Boosting Reasoning of Language Models Paper • 2502.04404 • Published Feb 6, 2025 • 25
