Teachometry
Run运行

Run · Get started运行 · 开始使用Node 24 · tutor-eval-v0.2a@0.2a.6

From research questions to reproducible runs.从研究问题走向可复现运行。

Run TutorBench locally, then connect an external Tutor to the Tutor Health workflow to diagnose concrete teaching failures in authored real-world scenarios. The canonical benchmark and frozen-evidence paths remain available as the underlying reproducibility layer.

先在本地运行 TutorBench,再把外部 Tutor 接入 Tutor Health 工作流,诊断人工编写真实场景中的具体教学失败。标准基准与冻结证据路径继续作为底层可复现层保留。

Run locally now立即在本地运行 View the documentation查看文档

  • Open source开源
  • Transparent evaluation透明评测
  • Evidence-first证据优先
Four glass test tubes of clear water stand in an aluminium rack on white paper, lit by one low lamp so their long shadows fall toward the camera. Each tube carries a paper label: 01 to 04, a public case ID, and the dataset tutor-eval-v0.2a@0.2a.6.
Cover封面 · Same cases. Different Tutors. Comparable evidence.相同案例。不同 Tutor。可比较的证据。

Five ways to run五种运行方式

Same questions.
Clearer evidence.
同样的问题。
更清晰的证据。

Run workflows运行工作流

01 / 05

Get started (provider-free)开始使用(无需服务提供方)

Clone the repository and run the deterministic Quickstart.

克隆仓库并运行确定性 Quickstart。

tutorbench · quickstart

# 1. Clone the repository
git clone https://github.com/shuangyan123/tutorbench.git
cd tutorbench

# 2. Install dependencies
npm ci

# 3. Run the Quickstart (no API key)
npm run quickstart
  • Repository main · Node 24仓库 main · Node 24
  • Runs locally without an API key无需 API key 即可本地运行
  • Uses 4 fixed cases from tutor-eval-v0.1@0.1使用 tutor-eval-v0.1@0.1 中的 4 个固定案例
  • No Judge, no network connection无需 Judge,无需网络连接
  • Deterministic checks (no official score)确定性检查(不产生正式分数)
  • Not leaderboard eligible不具备排行榜资格

A development/smoke demonstration with no overall score; Quickstart does not weaken the canonical benchmark boundary.

这是不产生总分的开发/冒烟演示;Quickstart 不会削弱标准基准的边界。

Published 0.1.0 · Node 22已发布 0.1.0 · Node 22Historical CLI with Quickstart; Tutor Health requires the Node 24 repository.历史 CLI 包含 Quickstart;Tutor Health 需要 Node 24 仓库源码。

npm install tutor-benchmark@0.1.0npx --no-install tutorbench quickstart

Current workflows use the Node 24 repository source. Published npm 0.1.0 requires Node 22 and includes Quickstart, but has no Tutor Health CLI or API. Check release channels and API boundaries 当前工作流使用 Node 24 的仓库源码。已发布 npm 0.1.0 要求 Node 22,包含 Quickstart,但没有 Tutor Health CLI 或 API。查看发行渠道与 API 边界

How to run如何运行

Traceable runs.
For better research.
可追踪的运行。
服务于更好的研究。

  1. 1

    Install安装

    Choose Node 24 repository source or the historical Node 22 package.

    选择 Node 24 仓库源码,或历史 Node 22 软件包。

  2. 2

    Run运行

    Start with Quickstart; Tutor Health requires the repository build.

    先运行 Quickstart;Tutor Health 需要仓库构建。

  3. 3

    Inspect检查

    Inspect findings, evidence, coverage, and versioned artifacts before the next Tutor change.

    在下一次修改 Tutor 之前,检查发现、证据、覆盖情况与版本化产物。

Execution boundaries执行边界

What can you run now?现在可以运行什么?

Different workflows serve different purposes. Here is what is available, what requires additional components, and what remains outside the current public scope.

不同工作流服务于不同目的。这里说明当前可用的内容、需要额外组件的内容,以及仍不属于当前公开范围的内容。

Provider-free无需服务提供方

  • Quickstart demonstrationQuickstart 演示
  • Public case/data inspection公开案例/数据检查
  • Frozen-corpus validation and replay冻结语料验证与重放
  • Website and artifact workflows网站与产物工作流
  • Dry-run and request validationDry-run 与请求校验
Run locally now立即在本地运行

External Tutor / model外部 Tutor / 模型

  • HTTP Tutor adapter (POST /respond)HTTP Tutor 适配器(POST /respond)
  • Product Tutor collection产品 Tutor 采集
  • Canonical model collection标准模型采集
  • Explicit Judge providers (OpenAI Responses, DeepSeek, generic Chat Completions)显式 Judge 服务提供方(OpenAI Responses、DeepSeek、通用 Chat Completions)
  • Advanced offline analysis workflows高级离线分析工作流
See configuration guide查看配置指南

Not yet public尚未公开

  • Calibrated public model results校准后的公开模型结果
  • Human calibration and validation人工校准与验证
  • Statistical validation统计验证
  • Official leaderboard rankings正式排行榜
Learn about the roadmap了解路线图

Make the run legible让每次运行都清晰可读

Reproducibility checklist可复现性检查表

Record the following information with your results to make them easy to verify, replay, and compare.

请将以下信息与结果一起记录,方便验证、重放和比较。

  1. Dataset version数据集版本e.g., tutor-eval-v0.2a@0.2a.6例如,tutor-eval-v0.2a@0.2a.6
  2. Case or corpus identity案例或语料身份Case IDs or frozen corpus案例 ID 或冻结语料
  3. Prompt / generation specPrompt / 生成规格Prompt templates and parametersPrompt 模板与参数
  4. Provider and model snapshot服务提供方与模型快照Model name, version, and settings (if applicable)模型名称、版本和设置(如适用)
  5. Run metadata运行元数据Date, environment, configuration日期、环境、配置
  6. Evidence output证据输出Logs, traces, and evaluation results日志、追踪信息与评测结果

tutor-eval-v0.2a@0.2a.6

Structured cases
for transparent evaluation.
结构化案例
用于透明评测。

  • 01correct-answer-wrong-reasoning-001
  • 02correct-answer-wrong-reasoning-001-zh-CN
  • 03fraction-misconception-001
  • 04fraction-misconception-001-zh-CN
Same cases.
Different Tutors.
Comparable evidence.
相同案例。
不同 Tutor。
可比较的证据。