Rubric Graded Reasoning: Multi-Domain Tasks with Expert Rubrics

A multi-domain reasoning dataset built to improve frontier models by revealing their failures and turning expert grading into training signal. The dataset pairs self-contained tasks with weighted rubrics across three domains — Computer Science, Data Science, and Chemistry. It was created by TuringEnterprises and last updated on 2026-06-16.

Use Cases

• Benchmarking model reasoning failures based on expert-graded rubrics mentioned in the description
• Training models to improve multi-domain reasoning based on the provided training signals
• Evaluating model performance on tasks in Computer Science, Data Science, and Chemistry as described

Strengths

• Tasks are paired with weighted rubrics for structured evaluation
• Covers three distinct domains: Computer Science, Data Science, and Chemistry

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment

Provenance

Source

TuringEnterprises via Hugging Face

Collection Method

Likely constructed from expert-designed tasks and rubrics

Freshness

Last updated 2026-06-16 18:50:57