STEPQuant: When and Where Errors Matter in Delta-Rule Recurrent State Quantization Paper • 2609.38169 • Published Sep 29 • 75
LimiX-2: A Contextual Mechanism Network Towards General Structured-Data Intelligence Paper • 2609.17488 • Published Sep 15 • 235
TRACE: Rollout-Guided Quantization-Aware Training for FP4 Reinforcement Learning of MoE Language Models Paper • 2610.07767 • Published Oct 6 • 43
DART-SD: Diamond-topology Aware Retrieval and Tuning for Self-Distillation of Multi-Turn Tool-Calling Agents Paper • 2608.18524 • Published Aug 19 • 27
Code as Worlds: Agentic Discovery of Executable World Representations for Physical Reasoning Paper • 2608.27549 • Published Aug 27 • 47