OpenAI and DeepSeek’s reasoning LLMs have scored impressively on benchmarks that challenge humans, but how robust are their fundamentals? We test o1-mini & R1 on small-scale graph coloring problems and find limited reliability and signs of issues with non-linear reasoning.
(1/10)
@BexBookaholic Beautiful! My wife and I were just there in July during a Jane Austen Society of North America tour of southern England. We were struck by how slanted the Cobb is. It would be easy enough to fall off - even if you didn’t jump like Louisa!
@BruceBartlett It’s awful that CBS won’t be fact-checking. But I wouldn’t want to miss the chance to see Walz (who has the highest net favorability ratings of all the candidates) pummel Vance (who has the lowest). 🍿
@McCormickProf If you look at X today, you’ll see that her rant is being repudiated by a great many of us liberals. Please don’t use her uncharitable remarks as a reason to be uncharitable to the other side yourself. Wouldn’t it be better for us all to turn down the heat on the culture wars?
@emhoneyreads When I was young, there was a TV show where the kid who won a contest got 5 minutes to run thru Toys R Us with a shopping cart and grab any toys they wanted for free. You’re right, we should have something like that for bookstores! 📚
@sewistwrites I agree with everything that’s been said - knitting is totally fine as long as it doesn’t make noise. People can listen while they knit. I personally rub a worry stone during meetings - it can be good to have something to do with your hands.