XCR-Bench - CSI Prediction: leaderboard
Metric: Exact Match (x100; case-insensitive exact matches of the masked culture-specific items, summed per sentence and averaged over sentences). Source: arxiv.org. Saturation forecast: Estimated already saturated. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude 3.7 Sonnet | 46.76 |
| 2 | GPT-4o | 43.7 |
| 3 | Qwen 2.5 Max | 43.7 |
| 4 | Gemini 2.0 Flash | 42.94 |
| 5 | Llama 3.3 70B Instruct | 34.54 |
| 6 | OLMo 2 32B | 31.87 |
| 7 | DeepSeek R1 Distill Qwen 32B | 29.96 |
Interactive version: theaggregate.ai/benchmark?slug=xcr-bench-csi-prediction · How It Works · Data refreshed daily, snapshot 2026-09-25.