Benchmark

ECLeKTic

A multilingual closed-book question answering dataset that evaluates cross-lingual knowledge transfer in large language models across 12 languages, using…

Modality
text
Categories
reasoning, language
Openness
unknown
Source
llm_stats
Reported scores
8

Reported scores

llm_stats

ModelOrganizationReported valueReportedEvidence
Gemma 3 12BGoogle0.1032025-03-12evidence
Gemma 3 1BGoogle0.0142025-03-12evidence
Gemma 3 27BGoogle0.1672025-03-12evidence
Gemma 3 4BGoogle0.0462025-03-12evidence
Gemma 3n E2B InstructedGoogle0.0252025-06-26evidence
Gemma 3n E2B Instructed LiteRT (Preview)Google0.0252025-05-20evidence
Gemma 3n E4B InstructedGoogle0.192025-06-26evidence
Gemma 3n E4B Instructed LiteRT PreviewGoogle0.0192025-05-20evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.