TutorBench¶
Product or Programme · All products and programmes
Evidence · 4 claims (3 for, 1 mixed) · 3 studies (3 design),q2· 1 of 3 report an effect size · 4 claims rest on one study
Description¶
TutorBench is a source-grounded benchmark dataset created by Bingxi Zhao et al. for evaluating adaptive multi-turn tutoring and customized practice generation with simulated learner profiles.
Components¶
- Public TutorBench sample: 30 tutoring scenarios with 324 structured rubric items across three use-case families: The public TutorBench sample is a rubric-level resource containing "30 tutoring scenarios and 324 rubric items across three use-case families: adaptive explanation, assessment and feedback, and active learning." The article analyzes its rubric metadata with a deterministic label mapping to measure agency-dimension and agency-skill density, finding one representative criterion requires the tutor to "provide hint[s] ... by asking at least 2 guiding questions without completely showing next steps." (Junyi Yao et al. (2026))
Claims¶
- DeepTutor improves overall first-person interactive tutoring quality by 10.76% over a Naive Tutor baseline on TutorBench [+M]
- Human raters' preferences align with the LLM judge on TutorBench (Pearson r = 0.82 across metric-level win rates) [+M]
- In the public TutorBench sample, agency-oriented rubric criteria are most concentrated in the active-learning use case (agency-dimension density 0.720, agency-skill density 0.411) [+M]
- Hint-not-solution is the most challenging rubric dimension, with scores ranging from 6% to 81% across models [~W]
Related Products and Programmes¶
- AcademiClaw
- Multimodal OBE examination-answer dataset of 485 scanned submissions with criterion-level rubric marks
Key Sources¶
- Bingxi Zhao, Jiahao Zhang, Xubin Ren, Zirui Guo, Tianzhe Chu, Yi Ma, Chao Huang. (2026). DeepTutor: Towards Agentic Personalized Tutoring. Technical Report, The University of Hong Kong. https://arxiv.org/abs/2604.26962
- Junyi Yao, Zihao Zheng, and Baichuan Li. (2026). Beyond Helpfulness: A Teaching-over-Solving Diagnostic for Measuring Educational Impact in LLM Tutors. arXiv preprint. https://arxiv.org/abs/2606.16206
Page metadata
| Field | Value |
|---|---|
type |
product |
id |
tutorbench |
title |
TutorBench |
description |
TutorBench is a source-grounded benchmark dataset created by Bingxi Zhao et al. |
status |
review |
product_kind |
dataset |
generated |
by process:settle-candidates on 2026-10-10 |
trust |
unverified — no one has confirmed this page's content |
Frontmatter is the page's machine-readable record — see the Schema & Guide.