How to reproduce
A deadline of Friday evening was set for an agreement between the Pentagon and Anthropic. It’s not clear if Trump’s announcement of a phase-out will equate to more time for negotiation or if the government is truly moving forward with firing Anthropic by declaring it a supply chain risk. The government may also seek to compel Anthropic to agree to its terms through the Defense Production Act, according to the Times. The government may also choose another AI partner, like Elon Musk's Grok, but CIA officials believe that product is inferior to Anthropic's, the Times reports.
，详情可参考91视频
姜云涛以铁腕思路开出一剂猛药：砍低效、调结构、聚焦主业、重塑增长。是止血回稳，还是再造新高？
Even though my dataset is very small, I think it's sufficient to conclude that LLMs can't consistently reason. Also their reasoning performance gets worse as the SAT instance grows, which may be due to the context window becoming too large as the model reasoning progresses, and it gets harder to remember original clauses at the top of the context. A friend of mine made an observation that how complex SAT instances are similar to working with many rules in large codebases. As we add more rules, it gets more and more likely for LLMs to forget some of them, which can be insidious. Of course that doesn't mean LLMs are useless. They can be definitely useful without being able to reason, but due to lack of reasoning, we can't just write down the rules and expect that LLMs will always follow them. For critical requirements there needs to be some other process in place to ensure that these are met.