Chartography: A New Benchmark for Pro Charts
echen · x · 2026-07-17
Chartography is a new benchmark for professional chart understanding. The authors point out that existing tests like ChartQA and CharXiv are basically saturated—frontier models often score 80%–90% on bar, line, and pie charts. They shifted to chart types closer to daily professional scenarios to test if models can truly comprehend real-world visuals.
This benchmark covers much harder charts, such as Kaplan–Meier survival curves common in oncology, engineering contour maps, stock/trading candlesticks, Sankey diagrams, and Bode plots. The authors note that the best current models score less than 45% on this benchmark, indicating that professional chart understanding remains a significant weakness.
Related event: Chartography: A New Benchmark for Professional Chart Understanding(3 posts)→
More from Research
- Project APE finds verifier reliability drops when papers contain multiple errors — soumitrashukla9 · 2026-07-22
- Project APE says verifier costs fell about 90x in a year as Chinese open models lead — soumitrashukla9 · 2026-07-22
- OpenAI-linked paper says capability RL can make models more reward-seeking — MariusHobbhahn · 2026-07-22
- Project APE builds its verifier benchmark from 100 AI-written papers with injected errors — soumitrashukla9 · 2026-07-22
- Paper proposes a CRED taxonomy and benchmark to measure research-error detectors — soumitrashukla9 · 2026-07-22
- OpenAI says long-horizon models need safety and alignment checks across full action sequences — rhiever · 2026-07-22