Knowledge: A Benchmark for Evaluating Knowledge-Related Tasks

A dataset for evaluating knowledge-related tasks, created by MRMRbenchmark and last updated in December 2025. The evaluation code is implemented based on the MTEB framework and is available in a linked GitHub repository. The dataset description emphasizes strict compliance with copyright and licensing rules from the original data sources.

Use Cases

• Benchmarking retrieval models based on knowledge-related tasks mentioned in the description
• Evaluating the performance of language models on knowledge-based reasoning
• Assessing model compliance with copyright and licensing rules based on the dataset's annotation guidelines

Strengths

• Evaluation code is publicly available and implemented using the established MTEB framework
• Annotation guidelines explicitly address copyright and licensing compliance

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download
• Column-level documentation is absent; field semantics must be inferred after download
• Row count and file formats are unknown, which may limit suitability assessment

Provenance

Source

MRMRbenchmark

Collection Method

Likely compiled for benchmarking purposes, with annotations following specific copyright guidelines.

Time Range

null

Freshness

Last updated 2025-12-10 00:04:26

Geography

null