Trained transformer-based chess models to play like humans (including thinking time) [P]

“`html

• A British researcher has trained a set of deep learning models, specifically transformer-based systems, to simulate human-like gameplay in chess.
• The models were fine-tuned for different rating buckets from ~800 to 2500+, using nearly one year’s worth of Lichess data. They include three types of models: a move model, a thinking time model, and a win/loss probability model.

“`

Stay ahead of AI. Get the most important stories delivered to your inbox — no spam, no noise.

Please enable JavaScript in your browser to complete this form.

Name Email

Name *

First

Last

Email *

Submit