VIVID - Topic Classification: leaderboard
Metric: Accuracy (%; exact-match choice of one of seven semantic themes for each idiom or proverb, three-shot, LM Evaluation Harness). Source: arxiv.org. Saturation forecast: Around April 2027. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Flash | 53.5 |
| 2 | GPT-4o | 52.4 |
| 3 | Llama 4 Scout Base | 45.9 |
| 4 | Qwen 3 14B | 44.7 |
Interactive version: theaggregate.ai/benchmark?slug=vivid-topic-classification · How It Works · Data refreshed daily, snapshot 2026-09-26.