Self-Retrospection Distillation: Turning Post-hoc Experiences into Prior Foresight Paper • 2610.08077 • Published Oct 6 • 168
X-Tree: Tokenizing Reusable Experience for Efficient Agent Generalization Paper • 2609.32993 • Published Sep 26 • 54
RRSI: Regularized Recursive Self-Improvement of Agent Harnesses Paper • 2609.24972 • Published Sep 21 • 164
Repo-To-Skill: Distilling GitHub Repositories Into AI4AI Skills Paper • 2609.02749 • Published Sep 2 • 148
BrowseComp-Plus: A More Fair and Transparent Evaluation Benchmark of Deep-Research Agent Paper • 2508.06600 • Published Aug 8, 2025 • 45