RADLADS: Rapid Attention Distillation to Linear Attention Decoders at Scale Paper • 2505.03005 • Published 8 days ago • 27
Writing in the Margins: Better Inference Pattern for Long Context Retrieval Paper • 2408.14906 • Published Aug 27, 2024 • 142
GoldFinch: High Performance RWKV/Transformer Hybrid with Linear Pre-Fill and Extreme KV-Cache Compression Paper • 2407.12077 • Published Jul 16, 2024 • 57