We all have our own strategy when it comes to trying to solve daily word puzzles such as Connections.
But one group of researchers from the New York University Tandon School of Engineering decided to work out how artificial intelligence programs fared at solving the game.
“Large Language Models are becoming increasingly widespread, and investigating where they fail in the context of the Connections puzzle can reveal limitations in how they process semantic information,” Graham Todd, a Ph.D. student in the NYU Game Innovation Lab, said in a statement in May.
Researchers found in preliminary results that AI was still far from mastering the game.
“Our research confirms prior work showing this sort of ‘chain-of-thought’ prompting can make language models think in more structured ways,” said Timothy Merino, another Ph.D. student. “Asking the language models to reason about the tasks that they’re accomplishing helps them perform better.”
How to Play ‘Connections’
The categories vary in difficulty from …