arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2610.14495cs.AIcs.GRcs.MMcs.SE
收藏

Reels-RSI:面向代码渲染解释器视频的程序级自我改进

Reels-RSI: Program-Level Self-Improvement for Code-Rendered Explainer Videos

Wei Lin

AI总结:

Reels-RSI是面向代码渲染解释器视频的开源程序级自我改进导演工具,通过规则学习与人为审核实现自我优化,经111个视频实验验证其缺陷捕获与防护效果,相关资源公开。

摘要:

解释器视频(如几何题解题过程、天空为何呈蓝色、王朝覆灭原因等)通常是像素视频生成的又一目标。本文认为它们是一种不同类型的产物:一种由确定性渲染器转换为帧的程序(脚本、分镜和逐帧动画代码)。程序可被读取、比对、测试和审计,而像素只能被观看。Reels-RSI是一款面向任何智能体框架的开源视频导演工具,其自我改进发生在程序层面,而非权重或提示层面。每一次修订过程都会记录代码快照;回溯环节将每个检查发现的问题与修复该问题的比对结果配对,并提出经验教训;对于静态检查可捕获的经验教训,会编译为带有必失败和必通过示例的规则,使记忆具备单元测试能力;且今日学到的规则会对之前生成的所有内容生效。受边界和基准约束的步骤允许智能体重写导演的剧本,仅当对保留主题的盲选成对判断更优时才会保留该剧本;没有任何内容会在无人为决策的情况下进入记忆。在111个视频(包含实地研究、两次剧本演进、2×2记忆消融实验、较弱导演、48个视频的学习曲线及判断者可靠性研究)中,我们发现:实地学到的规则捕获了其存在前生成的帧中的4个潜在缺陷,而未捕获其存在后的任何缺陷;规则是强导演的审计工具(8个无防护视频,无违规),也是弱导演的防护工具(30个无防护违规,5个防护违规);经验教训缩短了修订周期,但不会复合或改变判断者;首次重写的11-1裁决是投票设计中的位置偏差(判断者62%-81%的概率选择第二个展示的视频),掩盖了用13个事实换取技巧的脚本。我们记录了偏好信号的4种失败模式、每种模式对应的检查及修复方案。代码、规则、经验教训和数据均已公开。

英文摘要:

Explainer videos -- a worked geometry problem, why the sky is blue, how a dynasty fell -- are usually one more target for pixel video generation. We argue they are a different kind of artifact: a program (script, storyboard and per-frame animation code) that a deterministic renderer turns into frames. A program can be read, diffed, tested and audited; pixels can only be watched. Reels-RSI is an open-source reel director for any agent harness; its self-improvement happens at the program level rather than in weights or prompts. Every revision pass is logged with a code snapshot; a retrospective pairs each check finding with the diff that fixed it and proposes lessons; a lesson that a static check can catch is compiled into a rule with a must-fail and a must-pass example, so the memory is unit-tested; and a rule learned today runs over everything made before it. A bounded, benchmark-gated step lets an agent rewrite the director's playbook, kept only when blind pairwise judgments on held-out topics prefer it; nothing enters memory without a human decision. Across 111 reels -- a field study, two playbook evolutions, a 2x2 memory ablation, a weaker director, a 48-reel learning curve and a judge-reliability study -- we find that the rule learned in the field caught four latent defects in frames authored before it existed and none after; that rules are an audit for a strong director (eight unguarded reels, no violation) and a guard for a weaker one (thirty violations unguarded versus five); that lessons shorten revision loops but do not compound or move the judge; and that the first rewrite's 11--1 verdict was a position bias in our vote design (judges pick the second-shown reel 62--81% of the time), masking scripts that traded thirteen facts for craft. We document four failure modes of the preference signal, which check caught each, and the fixes. Code, rules, lessons and data are public.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑