neuron-explorer
Steering-first neuron explorer for Qwen3.5-4B: push every MLP neuron, keep every rollout, browse the results
翻译准备中
个人 GitHub 账号下公开仓库的所有者可以登录并验证所有权,自行管理介绍,无需管理员批准。
登录并生成验证文件,在 30 分钟内将其提交到此仓库的默认分支。其他分支或分叉仓库中的文件无效。然后在下方检查文件,即可无需管理员批准直接管理项目。请保留文件以维持管理权限。
如需登录,将前往登录页面。登录后请回到此处,再次点击所选的验证按钮。
在项目文件夹中运行 /nomorevibe verify,即可验证所有权并自行管理介绍。
一分钟完成验证正在加载完整 README。 以下为已保存的摘录。
Motiv is a .NET library for building composable, explainable boolean logic — so you never lose the why behind a true or false.
The boolean type has a problem: once evaluated,
you lose all context about why the value is true or false.
This is known as the boolean blindness problem:
Motiv addresses this by preserving the structure of boolean expressions, so you can recover the underlying causes when you need them:
This takes a lambda expression tree (Expression >) and transforms it into a hierarchy of propositions that mirror the expression's logic.查看完整 README ↗目前无法加载完整 README。请通过原文链接阅读。 以下为已保存的摘录。
Motiv is a .NET library for building composable, explainable boolean logic — so you never lose the why behind a true or false.
The boolean type has a problem: once evaluated,
you lose all context about why the value is true or false.
This is known as the boolean blindness problem:
Motiv addresses this by preserving the structure of boolean expressions, so you can recover the underlying causes when you need them:
This takes a lambda expression tree (Expression >) and transforms it into a hierarchy of propositions that mirror the expression's logic.查看完整 README ↗