DVMap: leaderboard
Current Large Language Models (LLMs) typically rely on coarse-grained national labels for pluralistic value alignment. However, such macro-level supervision often obscures intra-country value heterogeneity, yielding a loose alignment. We argue that resolving this limitation requires shifting from national labels to multi-dimensional demographic constraints, which can identify groups with predictab
Metric: Accuracy (self-reported). Source: benchmarklist.com. Saturation forecast: Not forecast. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o | 48.5 | #343 |
| 2 | Qwen 3 Next 80B A3B | 47.6 | #320 |
| 3 | Llama 3.3 70B Instruct | 46.4 | #534 |
| 4 | GPT-4o Mini | 46.3 | #599 |
| 5 | Qwen 3 14B | 46.2 | #537 |
| 6 | Gemini 2.5 Flash Lite | 45.3 | #422 |
| 7 | DeepSeek V3.2 Exp | 45.1 | #234 |
| 8 | Claude 3.7 Sonnet | 26.9 | #249 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=dvmap · How It Works · Data refreshed daily, snapshot 2026-10-20.