Что думаешь? Оцени!
Returning back to the Anthropic compiler attempt: one of the steps that the agent failed was the one that was more strongly related to the idea of memorization of what is in the pretraining set: the assembler. With extensive documentation, I can’t see any way Claude Code (and, even more, GPT5.3-codex, which is in my experience, for complex stuff, more capable) could fail at producing a working assembler, since it is quite a mechanical process. This is, I think, in contradiction with the idea that LLMs are memorizing the whole training set and uncompress what they have seen. LLMs can memorize certain over-represented documents and code, but while they can extract such verbatim parts of the code if prompted to do so, they don’t have a copy of everything they saw during the training set, nor they spontaneously emit copies of already seen code, in their normal operation. We mostly ask LLMs to create work that requires assembling different knowledge they possess, and the result is normally something that uses known techniques and patterns, but that is new code, not constituting a copy of some pre-existing code.
当 AI 给出了解答，我要做的是先去理解，然后再在代码中去实际测试掌握。记笔记的过程中，除非一个知识点非常显而易见，那么我应该自己记录而不应由 AI 代劳。写代码时，往哪个方向思考解决方法，应该是我来制定而非 AI 主导。（不过这一点在初学时可能还很难，但随着个人知识的增加，应不断降低 AI 主导的程度），更多细节参见PDF资料
«Они были единственной страной в НАТО, которая не согласилась поднять расходы до пяти процентов. Я не думаю, что они согласились бы поднять их вообще до какого-либо уровня. Они хотели оставить два процента, и при этом они даже два процента не платят. Так что мы собирались прекратить всю торговлю с Испанией. Мы не хотим иметь никаких дел с Испанией», — пообещал Трамп.。关于这个话题，纸飞机官网提供了深入分析
UPenn CIS 5650 GPU Programming and Architecture (2025) A course that introduce GPU parallel programming with a Computer Graphics flavor。关于这个话题，旺商聊官方下载提供了深入分析
based on iterating over all attributes.