Training
AITraining: The compute-intensive process of adjusting a model's parameters by exposing it to large datasets so it learns to predict or generate.
Used in these stories
- Why AI keeps picking 17: We tested randomness across 12 models
- OpenAI shares a proposed Navier-Stokes proof as researchers debate credit
- DeepSeek plans at least 160,000 Huawei Ascend chips in Inner Mongolia
- GenAI.mil added ChatGPT and Grok on the same day, both accredited at Impact Level 5
- Adobe values a year of free Firefly for 27 million people in Saudi Arabia at over $4bn, and publishes no method
- Brussels has designated ChatGPT a very large online search engine, on OpenAI's own count of 159.1 million users
- Microduck's press kit asks reporters not to call it open-source hardware
- Ox Alpha was GLM-5.3-Flash, a model that did not exist publicly when we fingerprinted it
- Anonymous model Ox Alpha counts tokens exactly like GLM
- An Unmonitored Anthropic Agent Deleted Jobs Inside A Cluster Holding Sensitive Resources