DKSplit on EuroHPC: CharBERT and ByT5
In our previous update, we outlined a systematic search for a model with enough world knowledge to handle the cases that DKSplit cannot: multilingual compounds, brand portmanteaus, domain names that...
Read →Domain Name Insights
AI Domain Registration Trends: What 38 Million .com Domains Tell Us
Key findings: Domains containing "agent" tripled in weekly registration volume since early 2025. For-sale rates for AI keywords run 2x-5x above the market baseline. When OpenClaw rebranded on January 30,...
Read →From Domain Segmentation to Reading Domain Signals
Can a general-purpose open model assess brand threat without any task-specific training? We tested four model configurations on 25 public domain disputes and found that they already recognize many of the right signals. But their readings change when you change which facts are presented first.
Read →How to Read Domain Registration Trends Like a Market
Not every spike in newly registered domains is a real trend. This post explains the independence test behind DomainKits Keyword Trends: how different registrants, nameservers, timing, and usage patterns help separate real registration momentum from one-operator noise.
Read →What’s Happening with 29,000 Short .COM Domains? A WHOIS Data Analysis
Short .com domains are among the most sought-after digital assets. Single-letter names like x.com, two-letter codes like ai.com ($70M), and three-letter names like NAS.com ($1.25M) command premium prices and sit...
Read →DKSplit on EuroHPC: Searching for a Teacher Model Across Architectures
Midterm report from EuroHPC Leonardo. We tested models across four architecture families. Each failed differently, and the search is narrowing.
Read →DKSplit Update: Cleaner Benchmark, First DeBERTa Run, Different Failure Modes
Continuing the work we started in our previous post. What We Did These Two Weeks Leonardo was in a planned maintenance window for most of the past two weeks. With...
Read →