Skip to content

shashwat-ghatiwala/pangram-perplexity

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 

Repository files navigation

Bee-wildered: ChatGPT-4o-mini struggles with the NYT Spelling Bee

I have been playing the NYT games for a few years now (Letterboxed, Spelling Bee, Connections, Wordle, etc.). I started off being horrible, but over time, I've gotten familiar with the different solves.

My favourite part of any game is playing it with someone - seeing how they do, teasing losses and celebrating wins. I was curious to see how good of a player ChatGPT-4o-mini would be. I decided to test its performance at the Spelling Bee. I've seen some people prompt engineer to get ChatGPT to solve the Wordle (some have tried with agents too).

For this project, I focused exclusively on the NYT Spelling Bee. More specifically, on one key aspect of the Spelling Bee -- the pangram i.e. the word(s) that gives you the most points.

About

ChatGPT-4o-mini fails at NYT Spelling Bee

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published