CSEBench¶
Product or Programme · All products and programmes
Evidence · 4 claims (2 for, 2 mixed) · 3 studies (2 design, 1 associational),q2–q3· 0 of 3 report an effect size · 4 claims rest on one study
Description¶
CSEBench is a publicly released benchmark dataset of 624 expert-annotated multiple-choice English-proficiency items with human response data, developed by Shangchao Min et al. for evaluating performance against China’s Standards of English Language Ability.
Components¶
Claims¶
- Closed-source LLMs reach CSE Level 6 while most open-source baselines cluster at CSE Levels 3–4 on CSEBench [+M]
- CSEBench items show adequate psychometric quality, with satisfactory difficulty, discrimination, and reliability for both diagnostic forms [+M]
- Benchmark success on human-designed assessments may not establish that the underlying abilities are being measured [~W]
- LLMs perform competently at comparing item difficulty but fall below random chance on comparing item discrimination [~W]
Related Products and Programmes¶
Key Sources¶
- Shangchao Min, Shaonan Wang, Xinyu Gao, Hui Wang, Zhiling Jin, and 2 more. (2026). Standardized Assessment of LLM English Proficiency. https://doi.org/10.21203/rs.3.rs-8820245/v1
Page metadata
| Field | Value |
|---|---|
type |
product |
id |
csebench |
title |
CSEBench |
description |
CSEBench is a publicly released benchmark dataset of 624 expert-annotated multiple-choice English-proficiency items with human response data, developed by Shangchao Min et al. |
status |
review |
product_kind |
dataset |
generated |
by process:settle-candidates on 2026-10-11 |
trust |
unverified — no one has confirmed this page's content |
Frontmatter is the page's machine-readable record — see the Schema & Guide.