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。
一分钟完成验证对 AI 代理进行支出控制和计费。使用 3 行代码防止 API 费用冲击。支持 OpenAI、Claude 和 MCP。
AI 翻译
Spend control & billing for AI agents. Prevent API bill shock with 3 lines of code. Supports OpenAI, Claude & MCP.
取自仓库 README
正在加载完整 README。 以下为已保存的摘录。
One spend ceiling for one agent job, consulted before each call. Bound to a task_ref you pass, not to a calendar month, a project or an API key.
A long agent run is many calls, often across several processes and more than one provider. A cap bound to a project, an organization or a calendar month is measured over that period. A per-call limit only ever sees one call. Neither of them is bound to the job.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。
One spend ceiling for one agent job, consulted before each call. Bound to a task_ref you pass, not to a calendar month, a project or an API key.
A long agent run is many calls, often across several processes and more than one provider. A cap bound to a project, an organization or a calendar month is measured over that period. A per-call limit only ever sees one call. Neither of them is bound to the job.查看完整 README ↗