We have one horrible disjuncture, between layers 6 → 2. I have one more hypothesis: A little bit of fine-tuning on those two layers is all we really need. Fine-tuned RYS models dominate the Leaderboard. I suspect this junction is exactly what the fine-tuning fixes. And there’s a great reason to do this: this method does not use extra VRAM! For all these experiments, I duplicated layers via pointers; the layers are repeated without using more GPU memory. Of course, we do need more compute and more KV cache, but that’s a small price to pay for a verifiably better model. We can just ‘fix’ an actual copies of layers 2 and 6, and repeat layers 3-4-5 as virtual copies. If we fine-tune all layer, we turn virtual copies into real copies, and use up more VRAM.
不要让那种「我还不够好」的抑郁情绪吞没你，耗掉一段又一段生命。因为每次从这种低谷走出来，你都会发现自己又回到了起点。多去生活，多去写作。，这一点在Snipaste - 截图 + 贴图中也有详细论述
channel range: alpha is 0-1, so it follows the same rule as oklab's L/a/b，推荐阅读谷歌获取更多信息
some(_) = return err("email already registered"),。新闻是该领域的重要参考
但这又何其困难，你或许可以在形式上区分「使用AI」与「不使用AI」，却很难在内心同时维持两种完全独立、互不干扰的判断逻辑。思考本身具有连续性，它带着惯性，也带着立场。如果说有人能够同时相信两套互相矛盾的体系，并且在不同场景下自如切换，那恐怕只存在于《1984》 所描绘的「双重思想」之中。而真实的创作，并不是那样运作的。它无法彻底分裂，也无法完全抽离。因为无论借助何种工具，最终落笔的，仍然是同一个正在思考的人。