AGIEval Gaokao Biology: Chinese University Entrance Exam Questions for AI Benchmarking

AGIEval is a human-centric benchmark for evaluating foundation models. This dataset contains the Gaokao Biology subtask, processed from the AGIEval repository. The data was authored by 'hails' and last updated on January 26, 2024.

Use Cases

• Benchmarking AI model performance on standardized biology exams based on the Gaokao subtask.
• Analyzing the difficulty and structure of Chinese university entrance exam questions for AI evaluation.
• Training or fine-tuning models on domain-specific educational content from a major international benchmark.

Strengths

• Derived from the established AGIEval benchmark, which is cited in academic literature.
• Specifically focuses on a high-stakes, standardized educational assessment (Gaokao).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Last updated 2024-01-26 18:36:41; freshness should be verified.

Provenance

Source

AGIEval repository on GitHub (Microsoft).

Collection Method

Processed as in the AGIEval repository, following existing dataset patterns on the Hugging Face Hub.

Freshness

Last updated 2024-01-26 18:36:41.