Comment by freediver

Impressed by the model so far. As far as independent testing goes, it is topping our leaderboard for chess puzzle solving by a wide margin now:

https://github.com/kagisearch/llm-chess-puzzles?tab=readme-o...