ChartCF achieves strong chart understanding performance in VLMs using significantly less training data by generating code-based counterfactuals, selecting similar samples, and performing multimodal preference optimization.
Start: Spatial and textual learning for chart understanding
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2representative citing papers
ChartVerse uses Rollout Posterior Entropy and truth-anchored inverse QA synthesis to produce 640K high-quality chart reasoning samples, training an 8B model that surpasses its 30B teacher.
citing papers explorer
-
Learning More from Less: Exploiting Counterfactuals for Data-Efficient Chart Understanding
ChartCF achieves strong chart understanding performance in VLMs using significantly less training data by generating code-based counterfactuals, selecting similar samples, and performing multimodal preference optimization.