Looping Beyond Twice: A Scalable Recipe for Looped Mixture-of-Experts Paper • 2610.01153 • Published Oct 1 • 15
DepthBench: Measuring How Residual Connections Enable More Computational Depth Paper • 2609.32534 • Published Sep 26 • 31
FlashLoop: Fast and Memory-Efficient Looped Transformers via Lazy Updates Paper • 2609.29812 • Published Sep 24 • 3 • 1
FlashLoop: Fast and Memory-Efficient Looped Transformers via Lazy Updates Paper • 2609.29812 • Published Sep 24 • 3
AlphaQ: Calibration-Free Bit Allocation for Mixture-of-Experts Quantization Paper • 2606.04980 • Published Jun 3 • 2