The Mathematics Search Engine

39+ min ago   (581+ words) When exploring large language models, you will often come across terms such as base model, chat model and reasoning model. These models are each useful, and describe different aspects of a model: how it was trained, how it behaves and…...

20+ hour, 4+ min ago   (328+ words) 3D printing news News Lawrence Livermore’s AI Vision System Catches Direct Ink Writing Defects Mid-Print Direct ink writing is one of those 3D printing technologies that has a process that can vary significantly. It’s not very commercialized, and mostly reserved for research…...

1+ hour, 2+ min ago   (821+ words) Iterations in one early set ate about 350 000 tokens each. The only reason anyone could say that out loud is that every task carried an estimate and a recorded fact — estimate, status, incidents and budget on one thread, or the cost…...

1+ hour, 9+ min ago   (406+ words) We're building FamiliaSync, an offline-first family organizer where all data stays encrypted on your own devices. That constraint extends to the assistant: every query has to run on the phone, no cloud calls allowed. Here's what happened when we replaced…...

1+ hour, 35+ min ago   (783+ words) Día 11 de la Serie de Arquitectura Open Source de Wisrovi. Al ejecutar grafos acíclicos dirigidos (DAGs) complejos en Python, los orquestadores tradicionales suelen estrellarse contra el Global Interpreter Lock (GIL) al correr transformaciones numéricas intensivas, o sufren un costo de…...

1+ hour, 37+ min ago   (1254+ words) My own code was steering a language model toward the same twenty popular games on one page in ten. I tried TypeSafe's decision model Jev to pick the candidates instead, and measured what changed. Tagged with ai, llm, jev, sideprojects....

2+ hour, 6+ min ago   (306+ words) When I started learning Python, I was mostly telling it what to print. Then I discovered if statements and loops. And suddenly, Python could make decisions, repeat tasks, and even help me analyse data. 👀 That was when things started getting…...

2+ day, 9+ hour ago   (122+ words) According to @_avichawla, MoE routing fragments batches, making grouped GEMM suboptimal at low concurrency and requiring weight-streamed GEMV kernels. Token routing to sparse experts splits the original batch into smaller expert-specific groups whose sizes depend on selection patterns, unlike uniform processing…...

2+ hour, 31+ min ago   (277+ words) The AI industry is moving incredibly fast. Every week we see better models, more capable agents, and new tools that automate increasingly complex work. But the more I study autonomous AI, the more I keep coming back to one question:…...

2+ hour, 54+ min ago   (296+ words) A copy-paste tutorial: generate a 1M-row CSV, describe the pipeline in four small YAML files, validate it before running, then compare the Python and Rust engines. Tagged with dataengineering, python, rust, tutorial....