Updated daily at 05:00 UTC
Benchmark
CharXiv (Reasoning)
Measures: Questions about 2,323 real charts taken from arXiv papers; the reasoning split requires combining information across several chart elements; scored by accuracy against reference answers.
- Maintainer
- Princeton Language and Intelligence, Princeton University (with UW-Madison and HKU)
- Introduced
- 2024
- Status
- Active
- Saturation
- Saturated
- Typical provenance
- Self-reported
- In scoring set
- No
Leaderboard
Independent and third-party results only.
No independent or third-party results yet. Show all results