Sign in to view source links and access this dataset
Description
A repository of standardized test question sets for evaluating large language models. The suite includes GRE, GMAT, SAT, TOEFL, and IELTS style questions curated into JSON datasets. It was created by Luoxi0416 and last updated on May 15, 2025.
Use Cases
Benchmarking LLM accuracy on standardized test questions based on the provided exam sections.
Comparing model performance across different test types (e.g., GRE vs. SAT) using the cross-dataset aggregation scripts.
Measuring inference latency and generating accuracy reports using the included Python driver and runtime summarizer.
Strengths
Includes a suite of five major standardized exams (GRE, GMAT, SAT, TOEFL, IELTS).
Provides utility scripts for cross-dataset accuracy aggregation and plotting.
Limitations
Column-level documentation is absent; field semantics must be inferred after download.
Row count is unknown, which may limit suitability assessment.
Provenance
Source
huggingface
Collection Method
Curated from exam-style question sets.
Freshness
Last updated 2025-05-15 07:36:47.
License is unknown; terms of use must be verified.