← All Posts
#LLM
#LLM
LLM
3 Posts
AI News
Granite 4.1 LLMs: A Look Under the Hood at How IBM Built Them
IBM's Granite 4.1 LLMs (3B, 8B, 30B) are trained on ~15T tokens with a five-phase...
AI News
Google’s TurboQuant Shrinks LLM Memory by 6x Without Sacrificing Quality
Google Research's TurboQuant compression algorithm cuts LLM memory usage by 6x and boosts speed 8x...
Deep Dives
Testing LLMs on Superconductivity Research Questions
Google researchers tested six LLMs on expert-level high-temperature superconductivity questions. NotebookLM and a custom system...