This one was a lot better than others. For every SAT problem with 10 variables and 200 clauses it was able to find a valid satisfying assignment. Therefore, I pushed it to test with 14 variables and 100 clauses, and it got half correct among 4 instances (See files with prefix formula14_ in here). Half correct sounds like a decent performance, but it is equivalent to random guessing.
[책의 향기]무기 팔고자 위협을 제조하는 美 군산복합체，推荐阅读服务器推荐获取更多信息
，推荐阅读搜狗输入法下载获取更多信息
Testing LLM reasoning abilities with SAT is not an original idea; there is a recent research that did a thorough testing with models such as GPT-4o and found that for hard enough problems, every model degrades to random guessing. But I couldn't find any research that used newer models like I used. It would be nice to see a more thorough testing done again with newer models.
Emer MoreauBusiness reporter，这一点在safew官方下载中也有详细论述
ВСУ запустили «Фламинго» вглубь России. В Москве заявили, что это британские ракеты с украинскими шильдиками16:45