Mainstream AI Agents in a Logic Number Puzzle Contest

AI agents keep getting better at math and reasoning, or do they?

I ran a straightforward and revealing test: how well do today’s mainstream AI agents solve Calcudoku puzzles?

I benchmarked 10 agents.
Results surprised me 👇
www.calcudoku.org/papers/ai_ag...

#AI #LLMs #AIEvaluation #Calcudoku

Puzzles Calcudoku en Ligne – Résolvez des énigmes numériques amusantes et intéressantes Jouez gratuitement à des puzzles Calcudoku de tous niveaux. Aiguisez votre logique et vos compétences en maths !

Version française de Calcudoku lancée ! Puzzles quotidiens gratuits.

😀🇫🇷

calcudoku.org/fr

Avis ?

#sudoku #puzzleslogiques #jeuxdelogique #calcudoku #Noël #Shein #PSG

A How to Solve video of the "best wishes for 2025" puzzle I posted a month ago 😎

www.youtube.com/watch?v=mBul...

#puzzle #puzzletime #coffeetime #saturday #indiedev #calcudoku #math #fbi #logicpuzzle #sudoku #followme

Finally a #dark mode at www.calcudoku.org 😀

🌜🌜🌜

#calcudoku #kenken #puzzle #indiedev

.. some tag surfing: #fbi #cybermonday #playstation #doj

2nd #bluesky post 😀

A surely doable 6x6 Calcudoku 😀

Each row and column has the numbers 1-6 exactly once.
Single cells are givens,
double cells also show an operation: when applied to the 2 numbers, it should produce the shown result.

#calcudoku #kenken #sudoku #logic #math
#logicpuzzle #puzzle