DishwasherBench – Can Frontier Models Stack a Dishwasher?
A dishwasher-loading game on a physical simulation of one real machine, built to benchmark language models and playable in a browser. The game inside is called Spotless. - composo-ai/DishwasherBench
Summary
A new browser-based game called Spotless challenges players to load a dishwasher using a physical simulation of a real machine. It aims to benchmark language models in an engaging way, blending gaming with AI testing.
Comments (39 of ) ⌛