VoteGEO-Bench
VoteGEO-Bench: Measuring the Impact of Generative Engine Optimization on Electoral Visibility
A benchmark measuring how webpage rewriting strategies affect candidate exposure in generative engine responses, comprising 28,200 query cells and 17,251 webpages from US and Japanese elections.
How to run
$ npm i -g lettokenburn-cli$ ltb bench info votegeo-bench
External harness. No Harbor adapter has been generated for this benchmark yet. `ltb bench info` prints the known links.
Runs happen on your machine with your own subscription; only the score summary is pushed. CLI docs
Scores
3 sourced entries| Harness | Source | Confidence | Link | |||
|---|---|---|---|---|---|---|
| GPT-4o-mini | 90.7 | N/A | Independent | medium | Oct 15, 2026 | arxiv.org |
| Gemini 2.5 Flash | 89.2 | N/A | Independent | medium | Oct 15, 2026 | arxiv.org |
| GPT-5-mini | 67.9 | N/A | Independent | medium | Oct 15, 2026 | arxiv.org |
Every number links to where it was published. Low-confidence rows are muted and need per-cell verification.
Reproduced by the community
All runsAbout
VoteGEO-Bench combines election-related queries, historical webpages, and voter personas from the 2024 U.S. presidential and 2026 Japanese House elections. It evaluates 11 rewriting strategies across four categories to assess their impact on target candidate visibility.
The study finds that target-specified rewriting increases exposure more than unspecified methods. Structure and Argumentation strategies yield stronger gains than Evidence or Style changes, with explicit partisan identities creating larger baseline exposure gaps than rewriting gains.