AI TCH JA Jayakrishnan M · Engineering with AI GLM-5.1: What a 754B Open-Source Model Topping SWE-Bench Pro Actually Means On April 7, Z.ai (formerly Zhipu AI) released GLM-5.1 under the MIT license. A 754-billion parameter mixture-of-experts model scored 58.4…
AI TCH SA Sam Hosseini · Paralleliq Beyond Prompt → Code: The Real Systems Challenges Behind Coding Foundation Models Most people think coding AI works like prompt in, program out. At the surface level, that is true. However, serious coding platforms…
