SAS: Simple Attention Sparsification via End-to-End Optimization of Context Ranking Paper • 2609.13141 • Published Sep 11 • 69
Attention Amnesia in Hybrid LLMs: When CoT Fine-Tuning Breaks Long-Range Recall, and How to Fix It Paper • 2606.11052 • Published Jun 9 • 17
Reasoning Arena: Trace Tournaments When Verifiable Rewards Fall Short Paper • 2606.09380 • Published Jun 8 • 11
The Depth Ceiling: On the Limits of Large Language Models in Discovering Latent Planning Paper • 2604.06427 • Published Apr 7 • 8
Believe Your Model: Distribution-Guided Confidence Calibration Paper • 2603.03872 • Published Mar 4 • 21
Efficient RLVR Training via Weighted Mutual Information Data Selection Paper • 2603.01907 • Published Mar 2 • 14