Install
The Mathematics Search Engine
Mathematics News & Resources
4Mathematics is a specialist search engine for Mathematics. Discover the latest math news and mathematical content. Part of the 4SEARCH network of topic specific search engines.
Latest Articles
Seven months of self-hosting our own AI stack: four bugs I can point at in the changelog
10+ min ago (376+ words) We run our team's AI stack on our own hardware. The weekly development log starts the week of January... Tagged with llm....
When an AI Agent Makes a Mistake in Production, Which Layer Should Stop It?
40+ min ago (1716+ words) A familiar production failure looks like this: an AI support agent reads a ticket, decides the customer deserves compensation, calls the refund tool, and refunds the full annual subscription instead of the $12 add-on. The model did not crash. The API…...
Why Most AI Agents Fail Long Before the Model Does
40+ min ago (1717+ words) The agent did not fail because the model was stupid. It failed because a CRM tool returned a 502, the agent retried, created two support tickets, read a stale knowledge-base article, filled the context window with stack traces, and then told…...
Ling 3.0 Flash Sante vs Llama Guard 4 12B - AI Model Comparison
2+ hour, 48+ min ago (102+ words) OpenRouter Ling 3.0 Flash Sante vs Llama Guard 4 12B: side-by-side summary Ling 3.0 Flash Sante and Llama Guard 4 12B are available through the OpenRouter API, so switching between them takes a model slug change rather than a new integration. Ling 3.0 Flash Sante, from inclusionai,…...
North Mini Code vs Qwen3.8 Max (0902) - AI Model Comparison
2+ hour, 6+ min ago (125+ words) Compare Qwen3.8 Max (0902) from Qwen to other AI models on key metrics including benchmarks, price, context length, and other model features. Access hundreds of AI models through the OpenRouter API. The latest top-tier model from major labs. Frequently chosen for programming…...
RAG Solved the Wrong Problem: What Actually Makes AI Applications Reliable?
40+ min ago (1652+ words) Your team ships an internal AI assistant grounded on company documents. The demo is excellent: ask about the refund policy, and the model answers with confident prose. Then production happens. A customer asks about a refund exception that lives in…...
Math Centers to Help Kids Learn
1+ week, 3+ day ago (784+ words) Pairing learning with activities and games is the best way to get kids to learn, and there are a lot of ways to make math centers to help kids learn numbers,......
Mathematicians made the world's fairest dice — they're 60-sided and can never end in a tie
1+ hour, 9+ min ago (883+ words) A team of mathematicians discovered that five 60-sided dice, arranged with exactly the right numbers, can determine a perfectly fair turn order for any group of players, with zero chance of a tie. There could be no ties and no…...
Kimi K3 vs Mistral Small 4: Benchmarks & Cost
1+ day, 12+ hour ago (312+ words) Supported · Public rank #8 Updated September 4, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload. Supported · Public rank #159 0 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and…...
Fermat's Last Theorem Machine-Checked: Claude Completes in 11 Days What Took Years to Plan
2+ hour, 30+ min ago (460+ words) That distinction — verification, not discovery — is important. It is also, for the long-term practice of mathematics, possibly more important than another discovery would have been. In 2024, Buzzard launched an EPSRC-funded five-year community project at Imperial College London to formalize FLT…...