AI 学习教练 · 人格驱动AI LEARNING COACH · PERSONALITY-DRIVEN

人格决定怎么教
证据决定教什么。
Personality shapes how we teach.
Evidence decides what to teach.

TraitTutor 是一位能长期对话、搜索和研究的 AI 学习教练。它用 Big Five 人格映射出你需要的支持方式,再用 BKT 学科证据追踪你真正掌握到哪里——每一次生成的学习页面,组件、顺序、节奏都为你而变。 TraitTutor is an AI learning coach that chats, searches, and researches with you over the long term. Your Big Five personality maps to the support you need, while BKT subject evidence tracks what you've truly mastered — every generated page adapts its components, order, and pace to you.

无需安装 · 浏览器打开即用 · 个性化可随时查看、修正或关闭No install · Runs in your browser · Personalization can be viewed, edited, or turned off anytime

支持画像 · BIG FIVESUPPORT PROFILE · BIG FIVE 组件编排第一输入First input to orchestration
尽责性Conscientiousness
82
开放性Openness
64
外向性Extraversion
45
宜人性Agreeableness
71
神经质Neuroticism
38

→ 映射为 SRL 支持维度,决定下一张学习页面→ Mapped to SRL support dimensions, shaping your next page

生成组件计划Generate a component plan
起点诊断Diagnostic 核心讲解Core Lesson 分步例题Worked Example 引导练习Guided Practice 校准复盘Calibration Review 主动回忆Active Recall 反思提示Reflection

很多 AI 工具记得你问过什么
却不知道你真正学会了什么
Many AI tools remember what you asked
but not what you've actually learned.

没有清晰的学习起点No clear starting point AI 可以解释知识,却不知道你的掌握程度,不知道从哪里开始、下一步学什么。AI can explain knowledge, but doesn't know your mastery — where to begin, or what comes next.
反馈只停留在当下Feedback ends in the moment 作答结果不会写进学习进度,后续讲解和练习也不会随之调整。Your answers never update your progress, so later explanations and practice never adapt.
下次还要重新交代背景Re-explaining yourself, every time 工具保存了内容,却没有保存学习——个性化停留在语气与难度,没有落实到页面本身。Tools save your content, not your learning — personalization stops at tone and difficulty, never reaching the page itself.

CORE · 人格个性化CORE · PERSONALITY PERSONALIZATION

组件编排的第一输入,
是你的人格。
The first input to component
orchestration is your personality.

Big Five / TIPI 先映射为 SRL 支持维度——支架强度、节奏、反馈、结构、挑战、互动——再叠加 BKT 学科证据与你的明确偏好,由组件选择器输出「选哪些组件、什么顺序、是否必须」。人格不是附加的语气层,而是每一张学习页面的编排依据。 Big Five / TIPI first maps to SRL support dimensions — scaffolding, pace, feedback, structure, challenge, interactivity — then combines with BKT subject evidence and your explicit preferences, so the component selector decides which components appear, in what order, and whether they're required. Personality isn't a tone layer on top; it's the orchestration basis of every learning page.

Big Five 人格Big FiveTIPI 可选量表Optional TIPI scale
SRL 支持维度SRL Dimensions支架 · 节奏 · 反馈 · 挑战Scaffold · Pace · Feedback · Challenge
+
BKT 学科证据BKT Evidence掌握 · 错题 · 复习Mastery · Errors · Review
=
你的学习页面Your Learning Page组件 · 顺序 · 是否必须Components · Order · Necessity

试着拖动你的人格Try dragging your personality

看看同一份材料,会为你生成怎样不同的学习页面。See how the same material becomes a different page for you.

为你生成的页面Your generated page

高结构 · 稳节奏
支架强度Scaffolding
结构清晰Structure
挑战强度Challenge
反馈温和度Feedback
互动频率Interactivity
节奏切片Pacing
1起点诊断Diagnostic Check确认先备Prereqs
2核心讲解Core Lesson概念主线Concepts
3分步例题Worked Example支架Scaffold
4知识图解Visual Map多模态Multimodal
5引导练习Guided Practice提示阶梯Hint ladder
6主动回忆Active Recall挑战Challenge
7迁移挑战Transfer Challenge挑战Challenge
8反思提示Reflection复盘Review

实际产品中,组件计划同时叠加 BKT 掌握证据、材料模态与你的显式偏好。In the product, the plan also weighs BKT mastery evidence, material modality, and your explicit preferences.

PERSONALIZATION

个性化在产品里,到底改变了什么?What does personalization actually change?

不是换一个皮肤、调一种语气。同一份材料,不同的学习者会得到结构、节奏、反馈和挑战强度都不同的学习过程。Not a new skin or a different tone. The same material becomes a genuinely different learning process — in structure, pace, feedback, and challenge.

🧭

结构Structure

有人需要先看到清晰的路线和检查点,有人更适合先从例子进入概念。Some need a clear roadmap with checkpoints first; others learn best by entering through examples.

同一份讲义 → 路线图先行,或例题先行Same handout → roadmap first, or example first
⏱️

节奏Pace

需要更多支架时,内容拆得更细;已经熟悉时,减少不必要的停顿。When you need more scaffolding, content is sliced finer; when you're fluent, it skips the pauses.

同一个知识点 → 五步小切,或一步直达Same concept → five small steps, or one direct leap
💬

反馈Feedback

有人需要更温和、确定的反馈,有人更喜欢直接指出问题所在。Some need gentle, reassuring feedback; others prefer the problem pointed out directly.

同一次答错 → 鼓励式引导,或直给式辨析Same wrong answer → encouraging guidance, or direct analysis
🎯

挑战Challenge

系统可以先给线索、拆成步骤,也可以增加主动回忆和自我测试。The system can offer hints and steps first — or dial up active recall and self-testing.

同一轮练习 → 提示阶梯,或迁移挑战Same practice set → hint ladder, or transfer challenge

HOW IT WORKS

先看目标和证据,再决定怎么教Goals and evidence first, then teaching

系统不是直接生成答案,而是先理解任务、读取证据、编排组件,再让可靠的作答结果回流,改变下一次生成。The system doesn't just generate answers — it understands the task, reads the evidence, orchestrates components, then lets verified answers flow back to change what comes next.

1
开始学习Start Learning输入学习目标、上传材料或直接提问Set a goal, upload materials, or just ask
2
读取支持方式Support Profile人格映射的 SRL 偏好与显式设置SRL preferences mapped from personality
3
读取学习证据Learning EvidenceBKT 掌握、错题、复习到期项BKT mastery, errors, due reviews
4
编排教学OrchestratePlanner 分派任务,生成组件 DAGPlanner dispatches tasks, builds the DAG
5
生成并校验Generate & Verify组件白名单渲染,Evaluator 把关Whitelisted components, Evaluator checks
6
作答回流Answers Flow Back服务端判分,更新状态与下一页Server-side grading updates the next page

PRODUCT

五个入口,一套持续生长的学习状态Five entries, one living learning state

对话、研究、学习路径共享同一份经过授权和裁剪的上下文——但它们都不会把「问过、读过、搜过」当成「已经掌握」。Chat, research, and learning paths share the same authorized, scoped context — but none of them mistakes "asked, read, or searched" for "mastered".

LEARN · 学习入口LEARN

上传材料,得到一条能完成的学习路径Upload materials, get a path you can finish

PDF、PPT、Word、图片或文本统一解析,建立一条可以持续推进、有历史记录的学习路径。PDF, PPT, Word, images, or text are parsed into a learning path that persists and progresses with you.

RESEARCH · 研究入口RESEARCH

复杂问题,先澄清范围再持续研究Complex questions: scope first, research deep

保存来源、主张、笔记与带引用的报告;研究结果还能继续转成课件、题目和复习材料。Sources, claims, notes, and cited reports are preserved — and results can turn into courseware, quizzes, and review materials.

MY LEARNING · 我的学习MY LEARNING

背景不用反复交代Never re-explain your background

目标、材料、进行中的路径和学习结果都在同一个空间里;从上次的位置继续,从薄弱点继续。Goals, materials, active paths, and outcomes live in one space. Resume where you left off — and from your weak points.

PERSONALITY · 性格设置PERSONALITY SETTINGS

完成 TIPI,或手动调整支持方式Take the TIPI, or tune support manually

结构、节奏、反馈和挑战强度随你调整。这些数据只调整学习支持,不用于判断能力或贴学习风格标签——你可以随时查看、修正、关闭或删除。Adjust structure, pace, feedback, and challenge your way. This data only tunes learning support — never judges ability or labels your learning style. View, correct, disable, or delete it anytime.

TOOLS · 学习工具LEARNING TOOLS

找到不会的那一步Find the exact step you're stuck on

输入题目或上传作业,系统结合 BKT 判断卡在概念、步骤还是应用,先给提示让你自己作答,再把错误变成后续的补充练习。Type a problem or upload homework. BKT pinpoints whether you're stuck on concept, steps, or application — hints come first, and mistakes become future practice.

BOUNDARIES · 边界与信任BOUNDARIES & TRUST

个性化有边界,证据有门槛Personalization with boundaries, evidence with thresholds

人格帮助你学得更好,而不是替你下结论。每一步都有明确的规则。Personality helps you learn better — it doesn't draw conclusions about you. Every step follows explicit rules.

01 · 教学诚信01 · INTEGRITY

不代写、不作弊No ghostwriting, no cheating 提供提示、讲解和练习;个性化不改变正确答案和判分标准。Hints, explanations, and practice only; personalization never changes correct answers or grading standards.

02 · 证据门槛02 · EVIDENCE

接触 ≠ 掌握Exposure ≠ mastery 问过、读过、搜过只算接触;只有可靠判分且归因明确的作答才更新 BKT。Asking, reading, and searching count as exposure; only reliably graded, well-attributed answers update BKT.

03 · 不做诊断03 · NO DIAGNOSIS

不贴人格标签No personality labels Big Five 只映射为可撤销的支持线索,不用于能力判断、学习风格定型或心理诊断。Big Five maps only to revocable support cues — never to ability judgments, fixed learning styles, or psychological diagnosis.

04 · 数据权利04 · YOUR DATA

可查看、可删除Your data, your call 用户数据相互隔离;学习记录与画像可以查看、更正、删除或关闭,基础功能不受影响。User data stays isolated; learning records and profiles can be viewed, corrected, deleted, or disabled — core features keep working.

GET STARTED

让学习结果,
真正改变你的下一步。
Let learning outcomes truly
change your next step.

答一道题,看状态如何变化,再看下一页如何为你调整——这就是 TraitTutor 与普通 AI 学习工具的区别。Answer a question, watch your state change, and see the next page adapt — that's the difference between TraitTutor and ordinary AI study tools.

免费试用 TraitTutor →Try TraitTutor Free →

注册即可体验完整系统 · 演示数据已脱敏Sign up to try the full system · Demo data anonymized