Sign in to view source links and access this dataset
Description
A dataset created by Anthropic, last updated on June 6, 2024, for evaluating language models. It contains questions designed to assess models' ability to handle election-related information accurately and harmlessly. The dataset is structured across three CSV files, each focusing on a specific aspect of election-related evaluations.
Use Cases
Benchmarking language model accuracy on election-related information-seeking questions.
Evaluating language model harmlessness in responses to election persuasion attempts.
Testing language model performance on European election-specific information.
Strengths
Created by Anthropic, a known AI research organization.
Structured into three distinct CSV files for specific evaluation aspects.
Last updated on June 6, 2024, indicating recent maintenance.
Limitations
Column-level documentation is absent; field semantics must be inferred after download.
Row count, file size, and sample data are unknown, which may limit suitability assessment.
Description metadata is limited; actual data quality requires manual inspection after download.
Provenance
Source
Anthropic
Collection Method
Likely contains evaluations implemented to assess language models.
Time Range
Creation or update timeframe implied by the 2024-06-06 timestamp.
Freshness
Last updated 2024-06-06 06:12:29
Geography
Includes a focus on European elections, as mentioned in the description.
License is unknown; terms of use must be verified before download.