-
TokDrift: When LLM Speaks in Subwords but Code Speaks in Grammar
Paper • 2510.14972 • Published • 35 -
LightMem: Lightweight and Efficient Memory-Augmented Generation
Paper • 2510.18866 • Published • 115 -
Every Attention Matters: An Efficient Hybrid Architecture for Long-Context Reasoning
Paper • 2510.19338 • Published • 117 -
The Smol Training Playbook
📚3.08kThe secrets to building world-class LLMs
Jonatan Borkowski
j14i
AI & ML interests
None yet
Recent Activity
upvoted an article about 13 hours ago
Welcome Gemma 4: Frontier multimodal intelligence on device liked a dataset about 16 hours ago
harborframework/terminal-bench-2.0 upvoted an article about 16 hours ago
Holo3: Breaking the Computer Use Frontier