docs(video-package): 新增 arXiv:2607.13104 §5 基座模型自我改进科普视频完整制作包; - #1100
Merged
Conversation
- 选题卡:3 标题方案 + 悬念钩子 + 受众/封面/标签建议 - 逐字稿:13:42 / 3868 字(4.7 字/秒自洽),N0–N6 段落 ID 主键体系 - 分镜表:46 镜头 11 列(md + csv 双格式),822s 与逐字稿/动画三表对齐 - 动画 demo:单文件 Canvas(1920×1080 / 8 场景 3B1B 风格 / 零依赖 file:// 直开,←/→ 空格 R 1-8 H 快捷键,sceneText 文案单一事实源) - 事实核查表:71 条断言全量回溯论文(引文 grep 100% 命中原文, RISKY/REWRITE 双零,ANALOGY 类比显式登记) - knowledge-map 登记制作包索引 🤖 Generated with [Claude Code](https://github.com/claude), [CodeX](https://openai.com), [Gemini](https://github.com/apps/gemini-code-assist) Co-Authored-By: Aurelius Huang<threefish.ai@gmail.com>
- S4 CUE_B:十张答案卡由统一「42」改为七张 42 + 三张杂项(41/38/44), 与分镜 S4-004 及旁白「十个里有七个相同」一致(others 数组从死文案转为实绘) - S1 CUE_A:公式注释「大脑记 θ · 装备记 Σ」缩短并下移至 y=368, 避开右上「工具箱」装备节点圆(圆底 y=342),实测包围盒零相交 - S3 CUE_A:阶梯注释上移至 y=900(实测底边 921 < 第一柱顶 940), 消除约 21 秒的文字压柱 - 28 处渲染层硬编码中文微标签全部收敛进 sceneText 各场景 glyphs 子对象, 兑现 README「文案单一事实源」承诺(区外硬编码中文已清零) - README 改文案说明同步标注 glyphs 微标签范围 验证:Chrome 实测三处修复点几何断言通过;全场景 0.5s 步长快进 OOB 复扫 0 越界;node --check 语法通过 🤖 Generated with [Claude Code](https://github.com/claude), [CodeX](https://openai.com), [Gemini](https://github.com/apps/gemini-code-assist) Co-Authored-By: Aurelius Huang<threefish.ai@gmail.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
背景
ThreeFish-AI/arxiv-paper-sci-video(commit4d8e671a)核心变更
01-topic-card/):3 个标题方案(主推《当 AI 开始给自己当老师》)、40 秒悬念钩子(「自己出题、自己批卷、自己验收——成绩单能信吗?」与保险丝段首尾呼应)、受众定位/封面文案/标签建议02-script/):13:42 / 3868 字,按 4.7 字/秒逐段自洽;N0–N6 段落 ID 体系为全包 join 主键;叙事将 §5 三条信号路径转译为「三种老师」心智模型——自己出题刷题(§5.1,模型坍缩风险)/ 自己批改作业(§5.2,confidently wrong)/ 亲自下场做实验(§5.3,世界模型脑内模拟),每条固定「怎么学 → 会不会学坏 → 怎么防」三段式,收尾落 §9.1 分层门控安全设计03-storyboard/):46 镜头 × 11 列,md 可读版 + CSV(UTF-8 BOM,可导入剪映/表格)双格式等价;822s 与逐字稿/动画三表精确对齐04-animation/index.html):单文件 Canvas,1920×1080 深色 3B1B 风格 8 场景;零外部依赖零网络请求,file:// 双击即开;←/→ 切场景、空格暂停、R 重放、1-8 直达、H 隐藏 HUD;全部中文文案集中于 sceneText 对象(文案单一事实源,改稿不动动画逻辑)05-fact-check/):71 条断言全量回溯论文——引文 100% 经 grep 验证命中论文原文、RISKY/REWRITE 双零、5 条 ANALOGY 类比显式登记「非论文内容」、8 个易错校准点(如 §9.1 实际标题为 Implications for System Design 而非「安全章」)逐项核实docs/.agents/knowledge-map.md登记制作包索引 1 条风险与回滚
4d8e671a)即可完整移除验证证据
影响范围
Next Best Action
video-package/04-animation/index.html通看 8 个场景确认视觉风格,满意后按 README 五步流程开录成片;文案微调只需改 sceneText 对象与逐字稿对应段落,三表按段落 ID 自动可追🤖 Generated with Claude Code