FreeToken: Efficient Edge-Native MoE Serving with Bandwidth-Adaptive Execution Paper • 2608.16157 • Published Aug 17 • 113
K-Search: LLM Kernel Generation via Co-Evolving Intrinsic World Model Paper • 2602.19128 • Published Feb 22 • 7
K-Search: LLM Kernel Generation via Co-Evolving Intrinsic World Model Paper • 2602.19128 • Published Feb 22 • 7
Sky-T1-7B Collection A series of 7B models trained with different recipes and the corresponding training data. • 8 items • Updated Feb 14, 2025 • 7
LLMs Can Easily Learn to Reason from Demonstrations Structure, not content, is what matters! Paper • 2502.07374 • Published Feb 11, 2025 • 41
LLMs Can Easily Learn to Reason from Demonstrations Structure, not content, is what matters! Paper • 2502.07374 • Published Feb 11, 2025 • 41