Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference Paper • 2412.13663 • Published 3 days ago • 82 • 8
Reducing the Footprint of Multi-Vector Retrieval with Minimal Performance Impact via Token Pooling Paper • 2409.14683 • Published Sep 23 • 9