Less is More: Recursive Reasoning with Tiny Networks
Paper
•
2510.04871
•
Published
•
446
read
-
-
Cache-to-Cache: Direct Semantic Communication Between Large Language Models
Paper • 2510.03215 • Published • 92 -
When Thoughts Meet Facts: Reusable Reasoning for Long-Context LMs
Paper • 2510.07499 • Published • 45 -
StreamingVLM: Real-Time Understanding for Infinite Video Streams
Paper • 2510.09608 • Published • 48 -
LiteStage: Latency-aware Layer Skipping for Multi-stage Reasoning
Paper • 2510.14211 • Published • 6 -
Every Attention Matters: An Efficient Hybrid Architecture for Long-Context Reasoning
Paper • 2510.19338 • Published • 92 -
LightMem: Lightweight and Efficient Memory-Augmented Generation
Paper • 2510.18866 • Published • 102 -
Glyph: Scaling Context Windows via Visual-Text Compression
Paper • 2510.17800 • Published • 57