NederlandsLanguage model
A language model is an AI model trained on vast amounts of text, so it can understand, write and translate text. The chatbots you know run on a language model.
A language model is an AI model that has learned from vast amounts of text. That lets it answer questions, write and summarise texts, translate and produce programming code. The well-known chatbots run on a language model. With a large language model, or LLM, both the model itself and the amount of text it was trained on are very large.
How does it work?
At its core, a language model learns to predict the next piece of text. It reads a passage, guesses what comes next and checks how far off it was. This happens billions of times. That is how it picks up patterns in language, but also in the knowledge contained in those texts. The text is chopped into small pieces, the tokens, and the model works with those pieces.
What the model has learned is stored in its parameters: billions of numbers. An answer usually emerges piece by piece, with each new piece depending on what is already there. Some models work differently. Mercury 3, from the American company Inception, first makes a rough version of the whole text and sharpens it in a few rounds. According to the company, that makes it four to eight times as fast as comparable models (link).
Why it matters
Language models are the engine behind almost all AI news. They form the basis of AI agents that carry out tasks on their own. They are also used in science, for example to design proteins or devise experiments.
Training such a model takes a lot of computing power, and a great deal of work goes into reducing that. Microsoft AI has finished the first training phase of its base model MAI-Base-2. According to the company, the model reaches the same level as a strong Chinese model with 4 to 12 times less computation (link). Researchers measure how good a language model is with fixed tests, the benchmarks.
Articles about Language model
-
AI model trained on almost four million fossils sheds new light on the great mass extinctions
Chinese researchers trained an AI model on almost four million fossils. It maps the history of life in unprecedented detail and suggests that three of the five great mass extinctions happened more gradually than thought. Other palaeontologists are somewhat more cautious.
-
A different order makes AI servers start answering more than three times faster
Engineers from Amazon and Mistral have made the open software vLLM, which many companies use to run AI models, work together more cleverly. With long questions, the first word of the answer appeared more than three times faster in their test. The new method is free to use.
-
Large open AI model from DeepSeek now fits on a computer with 128 gigabytes of memory
DeepSeek's new open AI model is too big for an ordinary computer. The company Unsloth made a heavily reduced version that runs on a powerful laptop or mini computer with 128 gigabytes of memory. This costs some accuracy.
-
Four AI agents finish Minecraft together using programs they wrote themselves
Four AI agents from Google DeepMind have finished Minecraft on their own. In just over 27 hours they shared out the work through chat, wrote 2,783 programs of their own and defeated the final boss. They played on easy mode, with software that tells them exactly what is around them.
-
AI model Mercury 3 writes more than a thousand tokens per second by creating the whole text at once
The American company Inception has released Mercury 3, a language model that does not write word by word but brings a whole text into focus in a few steps. According to the company, this makes it four to eight times as fast as comparable models.
-
Researcher says the inner workings of AI are much easier to understand than we thought
Nick Cammarata of the AI company Goodfire argues that the 'neurons' in AI models usually have one clear task. If that is right, it becomes much easier to look inside an AI model and check what it does. Other experts call the work interesting, but want to see more evidence first.
-
d-Matrix splits the work of an AI model across two kinds of chips for faster answers
At a major data centre conference, the American chip company d-Matrix showed how an AI model answers faster when ordinary graphics chips and its own chips each do the work they are good at. The company now also supplies complete computing clusters.
-
Chinese company Sugon shows computing cluster with 100,000 home-grown AI chips
China's Sugon is showing its Sugon 8000: a computing cluster that links 100,000 computing cards made in China. The system combines AI with classic supercomputing and, according to Sugon, has already run ten research projects.
-
Physicist lets AI swap calculation methods between fields: 36 papers in three months
Harvard physicist Matthew Schwartz had the AI model Claude apply calculation tricks from one field in another. With 19 colleagues, that produced 36 scientific papers in three months, from sunspots with a life expectancy to a rainforest that changes faster than expected.
-
Microsoft finds a more economical recipe for training large AI models
Microsoft AI has completed the first training phase of its new base model MAI-Base-2. According to the company, its new approach reaches the same level as a strong Chinese model with a quarter to a twelfth of the computing. The model is now being trained further.
-
Chinese company MiniMax releases fast and cheap AI model for agents
The Chinese AI company MiniMax has released M3.1 Flash, a model for AI agents that carry out tasks on their own. According to MiniMax, it is almost three times as fast as its predecessor and an agent task costs 70 per cent less. Independent tests will now follow.
-
Nvidia releases open AI models for digital assistants on servers and laptops
Nvidia has released three new open AI models for assistants that carry out tasks on their own. The largest runs in companies' data centres; the smaller ones simply work on a laptop. According to Nvidia, the large model beats top models on specialist tasks after extra training.
-
Does Claude learn by playing on a desert island?
MIT researchers let the AI assistant Claude loose on a virtual island for thirty hours without any task. It invented its own games, made drawings with blocks and did small experiments. With that experience, it later did better on most new tasks than a version of Claude without that play experience.
-
Made-up company data with a known answer teaches AI agents to analyse better
Researchers from Meituan and Chinese universities build fake worlds full of messy files, and they know exactly what lies behind them. That makes it possible to check whether an AI agent really draws the right conclusion. A model that practised with them became clearly better at data analysis.
-
Researchers let an AI agent think partly without words: faster and cheaper
Researchers from Stanford and Together AI let an AI agent do part of its thinking without writing it out. It only switches to text to look something up or to answer. According to them, it works up to 2.6 times faster this way and scores better than comparable systems.
