MemLoRA: Distilling Expert Adapters for On-Device Memory Systems Paper • 2512.04763 • Published 21 days ago • 3
VersatileFFN: Achieving Parameter Efficiency in LLMs via Adaptive Wide-and-Deep Reuse Paper • 2512.14531 • Published 9 days ago • 11
SonicMoE: Accelerating MoE with IO and Tile-aware Optimizations Paper • 2512.14080 • Published 9 days ago • 5
Improving Recursive Transformers with Mixture of LoRAs Paper • 2512.12880 • Published 11 days ago • 4